hash stringlengths 32 32 | doc_id stringlengths 7 13 | section stringlengths 3 121 | content stringlengths 0 2.2M |
|---|---|---|---|
d6e73de470353c70c32b5dcc77258e0b | 101 110 | 4.2.2 Examples of delays due to signal processing techniques | |
d6e73de470353c70c32b5dcc77258e0b | 101 110 | 4.2.2.1 Data produced by Matra Communication | In order to eliminate correctly the acoustic echoes perceived by the far-end listener due to the coupling between the handsfree loudspeaker and microphone placed in a car cockpit, the Acoustic Echo Cancellation (AEC) system needs to have efficient performances in terms of initial convergence, tracking in path variation situations, in adverse noisy and double-talk situations. Moreover the mobile network adds an important delay that will stress any bad or insufficient performances of the AEC if it is not well selected. For this purpose it is necessary that the AEC is mainly ‘full-duplex’ in any above-mentioned conversation situations. However existing component technology implies to define a minimal additional delay allowed for echo full-duplex processing and also for additive speech enhancement processing in noise. ETSI ETSI TR 101 110 V8.0.0 (2000-04) 24 (GSM 03.58 version 8.0.0 Release 1999) From simulations on real speech databases and realtime assessments, an additional parameter is defined, called Tadd_proc, taking into account any additional and mandatory speech processing block (including noise bad effects compensation) where Tadd_proc must is decomposed as follows: Delay for signal block size or sub-band decomposition: 16 to 24 ms Delay for noise reduction 12 to 16 ms Addition delay for computation 8 to 16 ms Tadd_proc 36 to 56 ms From this global Tadd_proc for speech processing block (possibly including noise reduction processing) it is also possible, and relevant in terms of extra-delay economy, to decompose the minimal requirement in three minimum values according to the kind of communication (Handset only or Handsfree). This intends to minimise the extra delays according to the kind of communications. These new decomposition leads to three kinds of Tadd_proc delays providing minimal values for acceptable echo or/and noise reduction and defined as follows: - Tadd_proc_AEC for Echo Cancellation when using Handsfree MS, Recommended minimum Tadd_proc_AEC: 28 (to 40) ms - Tadd_proc_NR when using Noise Reduction (NR) and a coupling reduction processing for use of handset MS, Recommended minimum Tadd_proc_NR: 20 (to 32) ms - Tadd_proc_HF for Hands Free when using AEC and NR when using Handsfree MS and if it is desired to add NR for listening comfort, Recommended minimum Tadd_proc_HF: 36 (to 56) ms It can be noticed that Tadd_proc_HF value is less than the sum of Tadd_proc_AEC and Tadd_proc_NR as the global processing will generally optimise the computational complexity for the association of two separate AEC and NR stages and consequently their global delay (Tadd_proc_HF). It is clear that these extra delays Tadd_proc_X could be strongly reduced, but the extra delay values defined above correspond to signal processing responding both to realistic computational constraints and to efficient performance assessed in actual conditions of GSM communications. To achieve a sensible extra processing delay reduction, by keeping the « full duplex » property, exponential- type increasing number of extra operations would be required: this is the case of RLS-based adaptive echo canceller working at normal full band and at sample. Another option to have a low extra delay is to use ‘Gain-switching’-based systems deploying low computational complexity. But the great disadvantages of such systems, in noisy and delayed transmissions contexts, are disastrous echo suppression and undesired switching of the useful speech to be transmitted. With such implementations double talk operation is not possible. |
d6e73de470353c70c32b5dcc77258e0b | 101 110 | 4.2.2.2 Data produced by Ericsson | For reasons on traffic safety, it is worthwhile to use a handsfree equipment while driving a car and using a mobile telephone. The use of handsfree equipment is today legislated in some countries, and legislation is on its way in may other countries. To obtain a good quality of conversation, the performance in this handsfree equipment should be as near full duplex as possible. The inherent long delay in today's GSM systems, max. 143.9 ms, makes the echo problem in handsfree situations much worse than in lower delay systems (e.g. ETACS, NMT, AMPS). The signal processing needed for these high quality full duplex handsfree solutions is extremely demanding. It requires a complexity which often exceeds that of today's speech and channel coding algorithms. Block processing has proven to be a successful way, in a consumer oriented digital signal processor (DSP), to cope with the high complexity. To fully exploit the benefits of block processing, it is important that the block length is sufficient: ETSI ETSI TR 101 110 V8.0.0 (2000-04) 25 (GSM 03.58 version 8.0.0 Release 1999) - to provide enough data for statistically good estimates of the properties of noisy speech; - to efficiently handle reverberation times in normal car cabins; - to provide adequate resolution in the frequency domain. It is also desirable, for reasons of efficiency in frequency domain processing, that the block length is a power of two. A block length of 256 samples, which corresponds to a 32 ms block at 8 ksamples/s, represents a good balance in these respects. With 32 ms blocks there will be an inherent delay of 32 ms. A reasonable figure for processing time of such blocks, using today's cost competitive DSPs, is 10 ms. Our proposal is to allow a minimum of 32 ms + 10 ms = 42 ms additional delay for handsfree signal processing. |
d6e73de470353c70c32b5dcc77258e0b | 101 110 | 4.3 Speech quality assessment | 4.3.1 General - Factors affecting the speech quality of the GSM system and derivatives Speech and Channel Coding Issues One of the most fundamental parts of the GSM system is the speech codec. To reach the required spectral efficiency, the speech codec is used to compress the speech data to minimise the amount of transmitted information. To achieve this, the speech codec effectively models the vocal tract of the user as a filter and sends the filter coefficients to the decoder, in the TRAU, together with some residual excitation energy. This basic method enables the speech codec to reduce the speech data rate to 13 kbit/s for the full-rate (TCH-FS) and EFR codecs and 5.6 kbit/s for the half-rate (TCH-HS) codec. The reduction in data rate adds distortion to the speech signal, the extreme effect of which can be heard in some of the military speech codecs, where the emphasis has been placed on intelligibility rather than speaker recognition. As a result the speech sounds 'Robotic'. The traditional measure for this distortion is the Quantization Distortion Unit or QDU. One QDU is equivalent to the amount of distortion introduced by a single transition from analogue speech to 64 kbit/s G.711 PCM and back to analogue speech. GSM Standard 06.10 states that the GSM speech codec introduces between 7 and 8 QDUs under error free (EPO) radio conditions. The QDU is an accurate measure for tandem PCM and ADPCM systems, allowing planners to determine the amount of distortion in a call routing path and ensure that it does not breach ITU-T recommended limits. ITU-T G.113 states that an international connection should not exceed 14 QDUs, this is broken down into the 5-4-5 rule. This allows the originating nation 5 QDUs, the international transit network 4 QDUs and the terminating nation 5 QDUs. Clearly a GSM to GSM call, national or international, will either just meet or breach these guidelines. Furthermore, there is some subjective evidence to suggest that the older generation, who grew up with an analogue PSTN equipped with electro-mechanical switches, will accept the 14 QDU limit. However the younger generation, who have only really known modern digitally switched fixed networks, will only accept an upper limit of 9 QDU. Another problem with speech codecs occurs when one codec is tandemed with another of either the same or a different type. Apart from the ITU-T G.721 32 kbit/s ADPCM standard, where synchronous coding adjustment allows tandeming to occur without any further distortion to be incurred, most calls will incur more distortion when speech codecs are tandemed. This is where the QDU begins to be an inaccurate measure as very low bit rate systems do not behave in a linear way. This is because some of the speech data, necessary for the second codec to produce an accurate representation of the input speech, has already been removed by the first codec, compounding the distortion effect. Although the ITU-T have now introduced the concept at present on the validity of these measures for planning the end to end distortion of connections with low bit rate systems. In addition, the standard only includes the Impairment Factors of ITU codecs. Tests would be required to assess the TCH-FS/HS and EFR codecs. Clearly, any improvement in speech quality is to be applauded. The recent adoption of the U.S.1. algorithm as the EFR codec has improved the speech quality of the GSM system, in error free or low error environments, to 'Wire-line' quality levels. The challenge now is to further improve on this advance to make it as invisible to the customer as possible that the telephone they are using is a cellular radio. ETSI ETSI TR 101 110 V8.0.0 (2000-04) 26 (GSM 03.58 version 8.0.0 Release 1999) Most cellular network operators are currently embarked on major cell build programmes aimed to provide high levels of coverage and capacity. Even with the large numbers of cells currently foreseen there are still areas of cells where the error performance of the system is poor. This is due to a number of factors. Radio planning tools rely on clutter databases to assess the building densities in a given geographic area. These databases do not hold a picture of the actual buildings but an approximation of what is actually present. This is also true for terrain databases. This means that the actual coverage differs from the predicted coverage. Drive testing can be used to check the quality and depth of coverage in a given area but this will not correct the entire network. The buildings that populate a given area all have different radio penetration losses due to the large variety of construction methods and even decor used. This cannot be planned for by the network operator. Hence, the occurrence of EP2 (C/I 7 dB) and EP3 (C/I 4 dB) is more frequent than desired. One method of overcoming these problems is to examine the possibilities of 'Robust' coding. There is already technology available that shows that a 'Robust' codec is possible and that can achieve 'Wire-line' quality across a broad range of operating conditions. In theory it could offer additional capacity in the network as the limit of C/I used for radio planning could be relaxed. In practice, should a 'Robust' codec be adopted, the SACCH signalling channel becomes the limit of performance. The SACCH is about 2-3 dB more tolerant to C/I than the current full-rate speech codec. An additional 2-3 dB of C/I margin would provide operators with additional fringe coverage and greater depth on in-building coverage. A major gain in radio capacity and coverage could be achieved if the SACCH performance could be improved to the same levels as a 'Robust' codec. This might provide an additional 5 dB or even 6 dB of C/I margin over the current full-rate system. The full benefits would not be realised until the majority of the operators customer base was equipped with 'Robust' mobiles, but the possibilities are worth exploring. Discontinuous Transmission (DTX) is another aspect of speech coding that affects the speech quality of the system. The Voice Activity Detector used to detect when the customer is speaking or not, inevitably introduces some clipping which can reduce the quality. Acoustic noise also effects the DTX system. The full-rate VAD was optimised to work with in car noise where the temporal characteristics are less dynamic than say street noise. As a result the VAD can be 'false' triggered, reducing the effectiveness of DTX at reducing C/I and extending mobile battery life. Frequency Hoping is useful to reduce the effect of fading on the speech path by hoping to another channel that is not affected by the fade. The complexity of a given speech codec has a knock on effect on the speech quality of the system. With a more complex algorithm, the speech codec can usually produce better speech quality for a given bit rate. However, to support that codec is mobile equipment, the signal processing technology has to be able to run the software in a reasonable period of time. Hence, complexity and delay are intrinsically linked. Delay, as we shall see later, can seriously degrade the quality of a connection. Terminal Issues Many, of the key speech quality parameters are determined solely within the GSM mobile. In addition, several of these parameters are critical of the interworking between the mobile network and any interconnected networks. The Send Loudness Rating (SLR), in the mobile to land direction, and the Receive Loudness Rating (RLR) in the land to mobile direction, determine the audio signal levels for the customers speech. The loudness ratings are calculated from the send and receive sensitivity masks or frequency responses. These are dictated by the acoustic transducers used as well as the anti-aliasing and reconstruction filters for the analogue to digital converters. One criticism levelled at GSM is that it is 'quiet'. This is difficult to understand as the SLR and RLR are in line with ITU-T long term values. It is important that the test method for TCL as well as the requirement are carefully considered. Echo problems were reported when GSM was first introduced, even on national calls. The GSM phase 1 TCL test allowed sinusoidal test stimuli to be passed through the speech codec, as some manufacturers use aspects of the speech codec in their acoustic echo cancellation devices. Unfortunately the GSM full-rate codec causes spectral spreading on sinusoidal signals which, when measured at their discrete frequencies, have a lower power than the equivalent speech signal. This meant that mobiles were passing the test but failing in an operational environment. The phase 2 test uses the ITU-T P.50 artificial voice to address this problem. Acoustic echo cancellation can be used as part of the TCL solution. However, to produce a stable acoustic echo canceller requires a complex algorithm due to the variable nature of the echo path. A complex algorithm requires additional delay which will cause other end to end speech quality problems. A low delay algorithm will not be as resilient but can be used as one part of the TCL solution. ETSI ETSI TR 101 110 V8.0.0 (2000-04) 27 (GSM 03.58 version 8.0.0 Release 1999) The interaction of acoustic noise on other system components can have a damaging effect on their ability to meet the needs of the customers. A particular example of this is the current half-rate codec noise problems where an acoustic noise rejection mask has been demonstrated to greatly improve the performance but has been rejected as it places stringent design restrictions on mobile terminals. An alternative algorithmic approach has been proposed but this will take as long, if not longer to implement, will require more delay and will still affect the speech quality. Noise coupling can have a serious affect of the public perception of the service offered. Handsfree terminals pose a particular problem to speech quality. The traditional vehicle mounted handsfree environment is particularly hostile. The acoustic volume is surrounded by a reflective material with high ambient noise levels. To overcome this the handsfree system has to use some algorithmic solutions but these introduce additional delay which causes problems with conversation quality. The recent adoption (ETS 300 903) of an additional delay for handsfree processing has increased the GSM one way delay budget by 40 % and is highly likely to cause problems with end to end quality. In addition a TCL limit (lower than 46 dB) has been adopted for handsfree. The argument for reducing the TCL is that the noise will mask any echo. However, the GSM handsfree car phone is installed in an environment where several people can use it. The system is usually installed for the benefit of the driver but a rear seat passenger may wish to use it, or the passengers may all wish to contribute to a call, requiring the volume to be increased so that they can all hear the system. To avoid future problems in the operational environment it would be better to further develop the test methods to use a Head And Torso Simulator (HATS) which more accurately represents the human that will use the terminal. In addition, carrying out a more system oriented set of tests using an artificial speech stimuli, such as the ITU-T P.50 algorithm, including the codec would be more appropriate. Network Issues The Mobile Switching Centre (MSC) incorporates an echo canceller adhering to ITU-T G.165 with at least a 60 ms echo path window. This is because the fixed network does not have echo control in the national network and the delay of the GSM system necessitates some form of echo control. It is important that the interaction of the MSC echo canceller with other network echo cancellers is understood. In an international connection between Europe and USA there will be an echo canceller in the home countries International Switching Centre (ISC) looking at the GSM network. This canceller will normally have a 64 ms echo path window but will be trying to cancel echoes over a 190 ms echo path. In addition the non-linear GSM speech codec renders the canceller ineffective. For this reason the ISC echo canceller should be switched out of the connection. The far end ISC will have an echo canceller looking at the far end customer. This canceller is connected in tandem with the MSC echo canceller. Tests carried out by BTL have shown that echo cancellers connected in tandem can reduce the amount of echo control by between 3 dB and 6 dB dependent on whether the centre clippers are active or not. For this reason the MSC echo canceller should be switched out. The mechanism for doing this is the CCITT n° 7 signalling system. By acting upon the information on the echo flag in the Initial Address Message (IAM) and the Final Address Message (FAM), the cancellers can be correctly controlled. In addition when choosing an echo canceller it important to realise that most cancellers have been built to meet the Blue Book G.165 which uses white noise to assess the cancellers performance. Unfortunately their performance can be reduced when operating with speech. The fixed links used to connect the BTS to the BSC and on to the MSC via the TRAU can affect the speech quality of the system. Fixed link errors can occur that can cause errors on the Abis speech frames which do not have any error correction on them. Ultimately, this could cause a bad frame to be seen as good by the TRAU. It has been suggested that TRAU bypass should be implemented to eliminate the effects of tandemed codecs. It must be remembered that this will only be true for connections between mobiles on the same coding scheme, e.g. full-rate. It should also be remembered that the same speech frame will be subjected to two radio paths. Hence, a mobile in an EP1 radio environment connected to another mobile in an EP1 environment could create an EP3 call. Codec tandeming is also a problem for non-real time communications. Voice messaging systems also use low bit rate speech coding to store the message. This adds another type of codec and more distortion to the connection. Call forwarding can also have an effect on the speech quality. Calls from a mobile to a fixed phone which has been diverted to a mobile will not be as good quality as a call which goes directly between the two mobiles as the echo control and delay will not be optimal. When developing interconnect agreements it is important to remember the routing of calls through the interconnected network. International calls, in particular, can be routed through Digital Circuit Multiplication Equipment (DCME) and Digital Speech Interpolation (DSI) systems. These add delay, distortion and clipping which may be unacceptable when combined with the GSM system. The international network also makes use of geostationary satellite routes which add 260 ms one way delay. GSM operators may wish to have their calls routed via cable connections as a preference. ETSI ETSI TR 101 110 V8.0.0 (2000-04) 28 (GSM 03.58 version 8.0.0 Release 1999) Interconnect Issues When developing and implementing a GSM mobile network it is important that the effects of the given design on the end-to-end speech quality are understood such that a design that offers acceptable quality across the full range of call scenarios, is implemented. This means that a balance must be found between the various engineering and commercial pressures that exist when designing the system. The difficulties of apportioning the various transmission parameters becomes increasingly difficult in a multi-operator environment. In many countries there is still only one fixed network operator with one or two cellular operators and a number of Private Branch Networks (PBNs). With this scenario it is relatively easy to apportion delay, loudness ratings and the other transmission quality parameters. However, as telecommunications liberalisation progresses most countries will develop a multi-operator environment similar to that in the United Kingdom where there are over 100 licensed operators, offering local, trunk and radio networks and services. Originally the United Kingdom relied on individual interconnect agreements between operators to agree the apportionment of transmission parameters across the network boundaries, guided by the Network Code Of Practice (NCOP). As the number of operators grew this technique became difficult to manage and the industry is now moving towards standard contracts with the transmission apportionment following the Network Performance Design Standards (NPDS) document produced for the Public Network Operators Interest Group (PNO-IG), a subgroup of the Department of Trade and Industry's (DTI) Network Interfaces Coordination Committee (NICC). The NPDS document uses a network model to apportion transmission parameters. End-to-end limits are taken from ITU- T Recommendations and individual system standards, such as the GSM Recommendations. It is assumed that there are local loop operators, Radio Local Loop (RLL) operators, such as Ionica, cellular operators, such as Cellnet, and trunk network operators, such as BT and Energis. One of the major principles of the document is that it is the responsibility of the operator that bills the customer, to deal with any complaints. This means that if a fixed customer calls a mobile customer, hears echo and subsequently complains to the fixed operator, the fixed operator has the responsibility to address the problem with the mobile operator. Unfortunately, in this example, the echo is usually due to the mobile customer's handset. The only safeguard that the mobile network operator has that the handset will not cause echo is that it has passed a type approval test, which we have already seen, can be inadequate and are under pressure to be removed. Current type approval test methods for digital cellular handsets do not fully represent their operational use and hence it is possible for handsets to pass the test but still cause echo whilst in use. The mobile operator has little or no control over the handsets connected to his network. The echo problems, if generated by a sufficient number of handsets and hence affect a large number of calls, can lead to litigation between the fixed and mobile operator. It is important that the type approval tests continue to be carried out and that they reflect the operational use of the handset. 4.3.2 Main Assessment Criteria for Handsfree processing used in GSM mobile environment From Field tests performed with handsfree GSM mobile phone in car and office environment it appeared that the main assessment criteria relevant to assess in a realistic way various handsfree processing solutions are the following (classified into two main categories): * Parameters affecting the Speech transmission quality: · clarity for users at each side of the communication · distortion on speech to be transmitted · Intelligibility of transmitted speech · Double-talk behaviour when users are talking simultaneously · clipping effect: on words end-points, hashing of words, generally characterising Echo suppresser techniques · fluctuation of the received voice at each side of the communication ETSI ETSI TR 101 110 V8.0.0 (2000-04) 29 (GSM 03.58 version 8.0.0 Release 1999) * Annoying artefacts due to the Handsfree processing: · residual echo level and nature and stability (e.g. convergence/tracking) · noise level · noise contrast effects The Handsfree processing stage can include an association of an Echo Canceller (generally equalisation adaptive filters), a Gain switching system and possibly a noise reduction module. 4.3.3 Evaluation Methodology for Full-Duplex Acoustic Echo Controllers developed within the FREETEL-Esprit project These data are derived from two publications referred as in reference [4] of subclause B.5 and reference [5] of subclause B.5. For assessing a Handsfree function of a mobile/fixed telephone a clear and reliable methodology of evaluation based on the speech quality objective criteria imposed on the Acoustic Echo Controllers (AEC) is needed in order to compare and assess several candidates of AEC. Hence an evaluation methodology, preferably designed in an objective way for reliability and costs reasons, must also include the definition of the test signals that may be speech or/and synthetic, and after that the selection of a reduced database corpus performed with the realistic acoustic front-ends used in the relevant environments (handsfree mobile phones in cars, offices...). An objective evaluation methodology for Acoustic Echo Cancellers (AEC) based on ITU- T recommendation G.167 has been developed within the FREETEL project1 and allowed to evaluate/compare the AEC algorithms performances on real echo/speech signals and acoustic front-ends recordings. However some modifications have been felt necessary in the project to adapt this objective methodology to the evaluation of a real Handsfree device with selected or given realtime integrated AEC algorithm in its context of application. One of the main interests of this real-context objective evaluation targeted by was to offer the possibility of correlating the objective criteria measurements with partial subjective tests performed by a speaker/listener using the handsfree device in the same time. In order to validate the proposed objective evaluation procedure for a real handsfree device it was led to dispose of a subjective evaluation methodology to assess the speech quality through the global handsfree system including: - the Acoustic Front-End (AFE) composed of the Handsfree terminal microphone(s) and loudspeaker(s); - the AEC algorithm with its, so-called, "glue" (corresponding to Double-Talk Detector (DTD), Voice Activity Detector (VAD); - the GSM network, which was in this case the Full-rate codec. Works on show, after first evaluations of AEC approaches developed in the project, some lacks of reliability on the objective criteria extracted from ITU-T recommendations for assessing handsfree functions of telephones, and, namely for GSM mobiles, adapted objective-subjective procedure based on the combination of a set of objective criteria measurements with limited subjective tests orientated towards the validation of the objective parameters is proposed. This evaluation methodology elaborated in FREETEL project is presented in two main parts: - Objective Evaluation procedure based on the parameters of ITU-T Recommendation G.167, including comments on the relevance of each parameter by the evaluation of Full-duplex AEC algorithms on real-contexts databases. This is described in next subclause 4.3.3.1. 1FREETEL : enhancement of handsFREE TELecommunications, project n°6166 funded by the ESPRIT III programme. ETSI ETSI TR 101 110 V8.0.0 (2000-04) 30 (GSM 03.58 version 8.0.0 Release 1999) - Adapted Objective Evaluation procedure based on the previous Objective Evaluation tool applied for the evaluation of any real-time echo canceller algorithm integrated in a real handsfree acoustic front-end device. This step of the Evaluation would be useful for the prototyping of a realtime echo canceller in a given handsfree acoustic front-end (AFE). This is described in next subclause 4.3.3.2. |
d6e73de470353c70c32b5dcc77258e0b | 101 110 | 4.3.3.1 Objective Evaluation procedure | An Objective Evaluation procedure is worthwhile if it complies with recommendations of telecommunications standards. A first and complete issue of ITU-T recommendation G.167 specifically applying Acoustic Echo Controllers (AEC) specifies a list of parameters to take into account for the working of an AEC placed in a given handsfree terminal for each kind of telecommunication networks, in particular for mobile radio systems. Recommended values for these parameters are generally provisional and are regularly amended according to the progress of the techniques and the real limits that would be given as feedback by handsfree telephones users: indeed it does not exist a clear specification of these limits, as also stressed in references [2], [3]of subclause B.5, however the recommended values provided by ITU- T Recommendation G.167 are precious and helpful for the AEC evaluation as they constitute the first and alone specifications with simple methodologies of tests provided in this context. |
d6e73de470353c70c32b5dcc77258e0b | 101 110 | 4.3.3.1.1 Objective Evaluation methodology of AEC devices | The Evaluation Methodology of AECs is based on a list of AEC performances parameters defined by ITU-T Recommendation G.167. This list, including 12 AEC performance parameters that acoustic echo controllers have to comply when placed in a handsfree terminal, can be classified into 3 main classes: - Terminal coupling loss parameters - Time adaptivity parameters Attenuation of speech in double-talk situations Sin Rout Sout Rin Echo Processing Block Network echo near-end speaker far-end speaker noise Figure 11: Handsfree audio terminal a) Terminal Coupling Loss Denoted TCL (or TCLw), is the overall attenuation of the echo resulting from the acoustic coupling of the terminal combined with the effect of the echo canceller. In fact it corresponds to the following ratio in dB, = ) ( ) ( log 10 Sout En Rin En TCL (1) ETSI ETSI TR 101 110 V8.0.0 (2000-04) 31 (GSM 03.58 version 8.0.0 Release 1999) Where En(.) defines the energy of respectively the Received input signal (Rin) and the Sent output signal (Sout) at ends of the audio terminal. In fact TCL (or TCLw) corresponds to the sum of two important quantities, the coupling loss between the loudspeaker and the microphone, denoted CL, and the Echo Return Loss Enhancement, denoted ERLE, which measures the intrinsic efficiency of the AEC algorithm. The coupling loss (CL) has the following expression: = ) ( ) ( log 10 Sin En Rout En CL (2) ERLE is the difference between the raw echo and the reduced echo got as output of the AEC processing whose expression is: = ) ( ) ( log 10 Sout En Sin En ERLE (3) with, TCL CL ERLE = + (4) if by assuming that Rin = Rout, i.e. no loss is present in the Receive way (case of an AEC). In the case of loss in the Receive way of the handsfree system - use of an echo suppresser (realised by gain switching) or more generally existence of attenuation/gain in the receive part - we must take into account this third quantity in the parameter TCL, the Receive Loss, denoted RL, = ) ( ) ( log 10 Rout En Rin En RL (5) which gives in the general case, TCL RL CL ERLE = + + (6) In the Evaluation Methodology it is taken into account more specifically the ERLE quantity (relative to the intrinsic performance of the AEC stage for a given acoustic front-end) as the TCL can be deduced from it by adding the RL and CL parameters that are generally considered as constant (RLcst, CLst) in the case of digital Handsfree functions. However for Echo Suppressers (ES) the RL quantity is not exactly constant but it can be noted that it is composed of a symmetric value of the ERLE plus a constant RL value. This can be summarised as follows: - for AECs: TCL = ERLE + (RLcst + CLcst) (7) - for ESs:TCL = 2.ERLE + (RLcst + CLcst) (8.1) where RL = ERLE + RLcst (8.2) From this, three main sub-quantities characterising the AEC in three specific situations as simple talk (st), double talk (dt) and echo path variation (pv) are defined: a.1. ERLEst/TCLwst which measures the terminal coupling loss in Simple Talk situation, i.e. the far-end subscriber is talking without near-end signal simulating the near-end speaker or the Handsfree terminal user. It is recognized that it is necessary to reach a minimum value of 46 dB in case of important network delays (e.g. GSM) and, on the other hand, it is really difficult to obtain such a high value without degrading the transmission quality, with the existing Echo Suppressers based on gain switching technology, in particular in transient situations as "dt" and gain switching phases in general. Indeed this deficiency of Echo Suppressers (Gain Switching) is moreover stressed by the important Networks delays. ETSI ETSI TR 101 110 V8.0.0 (2000-04) 32 (GSM 03.58 version 8.0.0 Release 1999) Now it is clear that such a limit for TCLwst is still considered as difficult and consequently in a full-duplex way as AEC algorithms cannot presently reach this value without unrealistic big computational load, which, then, precludes any integration of such complex algorithms on existing component technology fitting to manufacturing constraints of GSM terminals. It should seem more realistic to reduce this limit for handsfree terminals to take into account the real efficiency of AEC techniques that can be deployed with existing and cost-effective Signal Processing Components Technology. This was the case in the actual GSM Standard ETS 300 903. These figures represent a fair trade-off between maintained ‘Full-duplex’ properties (requested by users) and an acceptable level of residual echoes reduction. This parameter after successive evaluations was found as a reliable objective criteria pending that test signals are close to real speech signals. a.2. ERLEdt/TCLwdt corresponding to the previous case but with near-end speaker activity It is recognized that a TCL (and TCLw) limit in double talk situation could be lower than in single talk (Preliminary values are available in I-ETS 300 245-3). Evaluation Methodology implemented the proposed scheme and obtained results unable to separate correctly different AEC candidates. Other approaches to compute the ERLEdt were proposed in reference [4] of subclause B.5, but no clear conclusions were drawn, except that this parameter was not easy to represent. Hence it has been provisionally concluded that only subjective measurements could be performed. a.3. ERLEpv/TCLwpv measuring the echo loss during echo path variation: the AEC is let to converge to value ERLEst (or TCLwst) in the "st situation", then an echo path variation is applied during 5s, the ERLEpv is the ERLE measurement at the end of this time. TCLwpv should be greater than 10 dB. a.4. ERLEstn/ERLEpvn/ERLEdtn (TCLstn/TCLpvn/TCLdtn) measuring the echo loss during simple talk, echo path variation or Double-talk in noisy environments (i.e. when noise is added to the echo, or to the echo plus the near- end speech). The ERLExxn/TCLxxn evaluation procedures in noise have been kept similar to a.1, a.2, and a.3 procedures, but values here need to be defined. Moreover, depending on its spectral shape, noise may present different masking effects on the echo. So values must be defined for a given kind of noise to be selected among: white noise, low- frequency noise (car noise), medium frequencies noise (town & cars traffic) or speech-like noise (babble, cocktail party). In order to get reliable values of TCL in noise, better is to use noise with stationary spectrum like car noise or (filtered-) white noise. Even if values are not defined, the ERLExxn/TCLxxn evaluation in noise is recommended and allows to check and compare the robustness properties of different AEC techniques. b. Time adaptivity Parameters Time adaptivity parameters are intended to evaluate the performances of an AEC in different situations of communications. Five parameters have been identified in ITU-T Recommendation G.167 corresponding to four precise communication modes: b.1. Tic (initial convergence time) characterises the convergence behaviour of the AEC at the beginning of a communication assuming that the near-speaker does not talk or has already talked a first time. In ITU-T recommendation G.167 the echo attenuation in simple talk measured by the TCLst should be at least 20 dB after Tic=1s. This parameter has been integrated in the Evaluation Methodology and provided a quite reliable measurement pending that test signals are energetic enough on a sufficient time period and close to real speech signals (see Test Signals Database in subclause 4.3.3.1.2). b.2. Trdt (recovery time after double-talk) is the time necessary to recover a given echo attenuation after a double-talk event. The provisional limit in ITU-T Recommendation G.167 is Trdt=1s for a ERLEst of at least 20 dB. In practice the double-talk is generated during 2s after the AEC has reached TCLwst, then cut off and a timer is started to measure the re-convergence of the EC in simple-talk again (after double-talk event). This parameter was also integrated in the Evaluation Methodology but no conclusive results can be drawn. ETSI ETSI TR 101 110 V8.0.0 (2000-04) 33 (GSM 03.58 version 8.0.0 Release 1999) b.3. Trpv (time necessary to recover a given attenuation of the echo after an echo path time variation). Trpv should have the same limit as Trdt. The means to produce an echo path variation is still under study by ITU-T, so the "gain varying" technique applied on the echo present in the microphone channel was provisionally considered as a quite representative of an echo path variation in the Evaluation Methodology. b.4. Tonst/Tondt(break-in times in simple and double talk situations) are intended to measure the decreasing speed of transmitted speech attenuation. These parameters, computed for both Receive and Send sides, are more relevant to the Echo Suppressers case (half-duplex gain switching techniques). From Tonst, when attenuation becomes lower than 3 dB, Tonst should be < 20 ms. On the other hand Tondt should be < 20 ms when attenuation becomes lower than 6 dB. Hence in the Evaluation Methodology these parameters have not been retained as AEC techniques are full-duplex and no switching gain are applied on each channel. Nevertheless these quantities should be added if a full-duplex AEC (Adaptive filtering) is associated with an Echo Suppresser. b.5. Tic_n/Trdt_n/Trpv_n measuring the Time adaptavity or reactivity of the AEC techniques during simple talk, Double-talk or echo path variation in noisy environments (when noise is added to the echo, or to the echo plus the near- end speech). The corresponding evaluation procedures in noise have been kept similar to b.1, b.2, and b.3 procedures, but values here need to be defined and equivalent remarks on noise shapes to TCLxxn in a.4 are also applying. c) Speech Attenuation/Distortion in Double Talk mode These parameters are intended to measure the attenuation and distortion observed onto received and sent speech when double talk situations occur. These parameters are respectively denoted Ardt/Asdt and Drdt/Dsdt for receive/send speech attenuation and distortion. These parameters have not been used in the Evaluation Methodology as their interpretation has been found problematic for the case of AEC based on digital filtering (see subclause B.5 reference [4]). c.1. Ardt/Asdt correspond to the speech attenuation (received/sent speech) during double talk. A similar evaluation procedure as for TCLwdt can be used. Asdt/Ardt are obtained by comparing the speech just before adding 2s of double talk and the attenuated speech just after removing the double talk signal. In both sides, Asdt and Ardt should be less than 6 dB. c.2. Drdt/Dsdt correspond to the distortion of received/sent speech during double talk. The evaluation procedure is similar to the TCLwdt one, the distortion is obtained by comparing transmitted speech before and after double talk in terms of a distortion measure. Values are not specified and they are still under study. NOTE: For AECs Ardt and Drdt are not necessary since AECs do not alter speech signal in the receive way both for simple talk and double talk situations. |
d6e73de470353c70c32b5dcc77258e0b | 101 110 | 4.3.3.1.2 Test Signals used from the FREETEL-Esprit Database | Test Signals have been extracted from the FREETEL-Esprit database including handsfree telephone sets conditions for GSM applications in car and room (offices): handsfree acoustic part of the terminal was used for office echo recordings whereas separated microphone-loudspeaker(s) handsfree front-end was selected for echo recordings in car (Renault 25). In fact this last handsfree front-end corresponded to the commonly named "handsfree car-kit" now commercialised for most of the GSM mobile placed in car. Several kinds of signals have been recorded. These signal sequences include at the same time the far-end signal recorded at the loudspeaker input, (i.e. Rout) and the echo signal picked up at the microphone output (i.e. Sin): - real speech signals (1M/1F) for both English/French languages including couples of sentences and balanced energetic sentences for measuring AEC convergence/echo reduction performances such as (Tic, Trpv) and averaged (TCLst, TCLpv). ETSI ETSI TR 101 110 V8.0.0 (2000-04) 34 (GSM 03.58 version 8.0.0 Release 1999) - USASI signals which correspond to long-term average of speech signals and often used for their continuity of energy. They are very useful for tuning the AECs and also measuring (TCLst, TCLpv) parameters. - White noise sequences for the acoustic echo impulse response identification and for the analysis of transfer frequency function between the loudspeaker and the microphone. - CSS signals (Composite Source Signals) which are composed of high-, low-correlated signals and silence periods aiming at simulating speech signals. The use of real speech (couples of sentences & energetic sentences) or constant synthetic signals as USASI were preferred in this Evaluation Methodology. NOTE: This part of the FREETEL database is not presently available but a procedure to make it "public" is under way. |
d6e73de470353c70c32b5dcc77258e0b | 101 110 | 4.3.3.2 Adapted objective evaluation procedure to a GSM Handsfree mobile | |
d6e73de470353c70c32b5dcc77258e0b | 101 110 | 4.3.3.2.1 Motivation of an adapted evaluation procedure | In the context of the GSM mobile the objective methodology of evaluation presented in subclause 4.3.3.1 and in details in reference document [4] of subclause B.5, had to be focused onto the existing Handsfree device real-time platforms aiming at prototyping Industrial handsfree echo cancellation devices. The DSP board (EVM or commercial board), implementing the pre-selected AEC algorithms (after simulation), is connected through interfaces to the acoustic front- ends (the "Handsfree Adapter") and the PC-AT driving the application. The Handsfree Adapter allows to connect different acoustic front-ends such as a mobile handset or a ‘handsfree car-kit’ and permits to configurate the global system Acoustic front-end/PC-DSP in two possible ways of communications: internal mode or real GSM mode. On the other hand the Evaluation Methodology of AEC devices has been materialized into an Evaluation Tool (ET) implemented in C-language (see subclause B.5 reference [4]) based on each of the ITU-T G.167 parameters presented in subclause 4.3.3.1.1. However some modifications in the ET were felt necessary namely by only keeping objective parameters that were found reliable and relevant for AEC during preliminary evaluations on real acoustic front-ends signals for pre-selection of AEC algorithms. The separameters ??? are:: - echo reduction measures in Simple talk (‘st’) and Path Variation (‘pv’) modes: ERLEst/TCLst, ERLEpv/TCLpv, Tic, Trpv; - the similar parameters but computed in noisy conditions: ERLExxn/TCLxxn, Txx_n. The procedure is based on the use of the far-end terminal either in the handsfree or in the handset modes in order to take into account the possible distortion effects onto the far-end speech signal Rin going across the different communication stages inside the mobile terminal (A/D-D/A converting, speaker gain, speech encoding/decoding...) till the loudspeaker input Rout of the near-end terminal still in the handsfree mode. This procedure, presented in the following subclause 4.3.3.2.2, is materialised by recordings of useful signals for the echoes characterisation and AEC processing evaluation: the objective is to perform these recordings by using the global system PC-DSP-Handsfree Adapter above-described. On the other side the GSM network effect, which is essentially - the 90 ms delay making a double-delay shift onto the acoustical echo versus to the original signal, - the low-level noise signals, generated by the GSM network and that may perturbate the AEC filter, had to be included and, this, by performing a real GSM communication between the near-end and the far-end parts: this facility was offered by the Handsfree adapter that provided an interface with a GSM base. The following subclauses present this Adapted Evaluation procedure to be used to prototype any echo cancellation stage for a given acoustic front-end such as a handsfree car-kit (microphone and loudspeaker are independent) or a « self- contained handsfree » handset (the microphone and the loudspeaker are in the same box and close to each other). Then it is derived and proposed from this procedure a scheme for assessing a whole Handsfree device for GSM terminals. ETSI ETSI TR 101 110 V8.0.0 (2000-04) 35 (GSM 03.58 version 8.0.0 Release 1999) 4.3.3.2.2 Objective evaluation procedure for prototyping the GSM Handsfree AEC algorithms The proposed procedure was based on the use of a Head and Torso Simulator (HATS) successively placed in front of the different kinds of terminals connected to the Handsfree Adapter. The procedure decomposes in two main steps: a) creating of a specific handsfree speech database within the GSM mobile platform context b) objective evaluation of the AEC processing by using pre-recorded speech signals resulting from step a) The following block-diagram provides an illustration of the procedure tool. near-end HATS Evaluat. Tool (1) Sin Rout Sout Rin Lsp1 Mic1 Lsp2 far-end HATS(2) Mic2 Handsfree Adapter PC/EVM GSM Network Noise Figure 12: Block-diagram of the Evaluation system for a Handsfree Mobile/Fixed Terminal a) Pre-recording and storage of the specific handsfree speech database This phase corresponding to the pre-recording of handsfree speech materials will be only done once. It is composed of three steps, Step 1: Diffuse by means of HATS (2) the speech input files of the FREETEL -Esprit database (see subclause 4.3.3.1.2, loudspeaker input files) corresponding to speech/synthetic signals. In the same time these signals are recorded both at the Lsp1 loudspeaker input, denoted Rout, and at the Mic2 microphone output, denoted Rin, in a digital format and stored into the hard disk memory of the external PC driving the DSP board. Rin can be recorded either at the Handsfree adapter interface (input of the echo cancellation system), including the GSM network , or at the far-end Mic2 output. Recordings can be synchronised at the far-end with the near-end. Step 2: Newly stored signal Rout is then transformed according to a linear rule of gain varying for simulating the echo path variation. New transformed signal will be denoted Rout_pv. All these operations/computations are supported by the PC. Step 3: Signal Rout (or Rout_pv), stored in the PC memory, is diffused by the PC through the loudspeaker Lsp1, the Sin signal at the microphone Mic1 output is recorded and stored in the PC memory. For this step of the database recording two kinds of situation are foreseen: - in simple talk situation with the near-end HATS (1) in silent state, - in double talk with HATS (1) diffusing in the same time a near-end speech. - both situations with ambient noise generation extracted from a library of pre-defined noise files ETSI ETSI TR 101 110 V8.0.0 (2000-04) 36 (GSM 03.58 version 8.0.0 Release 1999) b. Objective evaluation of the AEC processing on the Handsfree Database After the specific database recording task realised once in a. the objective evaluation of AEC algorithms implemented into the DSP can be started every time one desires. During the AEC processing, the following operations will be done in the same time: - the pre-stored Sin signal (echo + possible near-end speech produced by HATS(1)) is being diffused; - the pre-stored Rin signal is being diffused; - the output signal from the AEC denoted Sout in the figure will be provisionally stored into the PC memory. We recall that the output signal Sout will represent in this case the residual echo signal + possible transformed "dt" signal and environment noise. After that the Evaluation Tool (ET) will provide by using all necessary stored signals the AEC performance parameters previously defined according to 4.3.3.1.1 and reference [4] of subclause B.5 and the deviation to the ITU-T recommended values. In fact the ET software will be preferably placed in the PC and will give these results under a synthetic table format every time one desires to test the AEC algorithms integrated inside the DSP board. Then these results can also be compared to the simulation results obtained with the same AEC methods implemented in C-unix. Nevertheless only the set of parameters retained in 4.3.3.2.1 are considered with reliance. This objective evaluation procedure will allow to check preferably the performances of the real time AEC versions with regards its simulated version |
d6e73de470353c70c32b5dcc77258e0b | 101 110 | 4.3.3.2.3 Proposed Objective-Subjective Evaluation procedure | The proposed Objective-Subjective evaluation consisted in performing listening tests by using the Handsfree system Objective ET platform presented in subclause 4.3.3.2.2. A limited number of expert listeners (e.g. 10 listeners) will give quality marks well targeted on the main handsfree effects to be analysed , previously defined in subclause 4.3.2 such as echo disturbance, bad tracking/convergence on the echo, clipping effects, transmitted ambient noise contrasts. The Handsfree system ET will compute in the same time the retained Objective parameters. This is intended to provide some feedback on the objective parameters thanks by the evaluation of subjective quality performances done by the expert listeners. This Objective-Subjective Evaluation procedure was not fully validated but was partly used for real implementation of AEC techniques on target DSPs. This might be used also to validate firstly-retained objective parameters (in 4.3.3.2.1) and make necessary modifications/refinements on those that are not easy to interpretation/implementation (double talk situations). If this first step is considered as satisfactory then an objective evaluation based on the pre-validated objective parameters could be only done for each handsfree function to be tested. From experiments performed with the FREETEL handsfree system in real GSM communications it has been felt mandatory to keep, for the Handsfree device evaluation, the GSM network component since it has a significant influence onto the AEC filtering stages. |
d6e73de470353c70c32b5dcc77258e0b | 101 110 | 4.3.3.2.4 Derived Evaluation schemes for Handsfree mobile telephones | From the previous objective evaluation procedure of Handsfree AEC processing devices it is also interesting to derive the same type of procedure for Handsfree GSM terminal products. This procedure permits to compute a set of 11 objective parameters retained in subclause 4.3.3.2.1 aiming at validating the main performances of the AEC device integrated in the Handsfree terminal. This set of 11 parameters is computed by means of the Evaluation Tool software mentioned in subclause 4.3.3.2.1 and are summarised hereafter: - Average echo loss (TCLst) and convergence times in simple talk Tic 20 (till TCL reaches TCLst -20dB) and Tic 10 (till TCL reaches TCLst - 10 dB). Tic characterises the reactivity of the AEC when the communication starts; - Average echo loss (TCLpv) and convergence times after echo path variation Tic_pv_10 / Tic_pv_20 (at TCLpv=10dB/20 dB) Tic_pv measures the reactivity of AEC after "path variation" of the acoustic echo; - Performances of echo reduction in noisy environments: (TCLstn, Tic_stn) and (TCLpvn,Tic_pvn); ETSI ETSI TR 101 110 V8.0.0 (2000-04) 37 (GSM 03.58 version 8.0.0 Release 1999) Noise is added on the acoustic echo and must present preferably a stationary spectrum; - Behaviour assessment of a purely "Half-duplex" device (gain switching) in interactive situations between the near-end and the far-end users: the Tonst parameter measure is recommended to be added but this parameter need to be analysed. The following schemes of evaluation can be used for measuring performances of AEC systems integrated in GSM terminals. Sin Rout Sout Rin Lsp1 Mic1 Evaluat. Tool PC/EVM Switch ADC DAC Gain- Echo Suppr. Full-duplex Adaptive Echo Canceller cod near-end signals in PC UPCMI MS/Car Kit Interfaces GSM Mobile Station (MS) dec (or HATS) (+noise) Figure 13: Evaluation of the AEC function of a GSM Terminal Same procedure can also be used for GSM terminals equipped of a Handsfree car-kit: ETSI ETSI TR 101 110 V8.0.0 (2000-04) 38 (GSM 03.58 version 8.0.0 Release 1999) Sin Rout Sout Rin Lsp1 Mic1 Evaluat. Tool PC/EVM Switch ADC DAC Gain- Echo Suppr. Full-duplex Adaptive Echo Canceller cod near-end signals in PC UPCMI decod (1) (2) External Handsfree car kit 2 options : (1) HF processing in MS (2) HF processing in car kit Analog I/O connected to MS UPCMI Interface : Connection to MS Digital I/O GSM Mobile Station (MS) MS/Car Kit Interfaces (or HATS) (+noise) Figure 14: Evaluation of the AEC function of a GSM Terminal with Handsfree Car-kit Remark: these both schemes set up the problem of the UPCMI connections to the inputs or outputs of the AEC processing modules integrated in the existing GSM terminals. In order to make the measurements applicable it is necessary to dispose of the Rin/Sout digital signals measures just before/after the encoding/decoding blocks. If the GSM terminal is equipped of a Handsfree car-kit, generally AEC processing is performed in the external box and it is easier to pick-up these signals. However in future terminals it is foreseeable to have all digital AEC processing within the MS itself, but in this case the signals measures will present same constraints as previously. ETSI ETSI TR 101 110 V8.0.0 (2000-04) 39 (GSM 03.58 version 8.0.0 Release 1999) |
d6e73de470353c70c32b5dcc77258e0b | 101 110 | 4.3.4 Subjective tests | |
d6e73de470353c70c32b5dcc77258e0b | 101 110 | 4.3.4.1 Subjective opinion tests | The approved subjective methods which are used to determine the quality performance of a given telephone connection are described in Recommendation P.800. Two categories of tests are identified: Conversation Opinion Tests and Listening Opinion Tests. Conversation Tests are intended to reproduce as far as possible in the laboratory situation the actual service conditions experienced by customers. They are mandatory to estimate the effect of degradation's linked to the two-way connection: talking degradation, e.g. sidetone and echo, and conversation degradation, e.g. propagation time, mutilation of speech by the action of voice operated devices; which are typically degradation's for handsfree telephone. Listening only tests or one-way listening tests have been extensively used when the degradation factors to be studied affect only the listening of customers. |
d6e73de470353c70c32b5dcc77258e0b | 101 110 | 4.3.4.1.1 Listening opinion tests | Several methods are recommended and fully described in Annexe B to Annex F of ITU-T Recommendation P.800. Among these methods, the two most frequently used are the ACR (Absolute Category Rating) and the DCR (Degradation Category Rating) methods which are both category-judgements methods. One collects the subject’ vote on a five-point Quality Scale (ACR), the other on a five-point Degradation Scale (DCR). For both procedures, a strict control of all experimental factors (from speech samples used for recordings to orders of presentation) through experimental design allows to obtain reliable results, which are, thus, statically validated. Naive or non-experienced subjects are used for the evaluation. |
d6e73de470353c70c32b5dcc77258e0b | 101 110 | 4.3.4.1.2 Conversation Opinion Tests | This procedure is an interactive evaluation procedure involving two participants: a talker and a listener. They are both placed in a controlled environment and invited to hold conversations with the help of a proposed conversation task. At the end of this exchange, they are individually asked to assess the perceived quality of the communication. Various 5- point category judgements are used; two opinion scales are recommended in P.800, Annex A: the classic Quality scale and a Difficulty scale (binary response to the question:"Did you or your partner have any difficulty in talking or hearing over the connection ?". Certain listening laboratories use more than two scales, typically a questionnaire with half a dozen items. CNET recommends a multicretiria approach where descriptive criteria are grouped together and the use of a minimum of a quality scale and of two impairment scales, namely an echo annoyance scale and a noise annoyance scale. The impairment scale is taken out the DCR method but its application to the conversation situation differs from the recommended procedure for the listening tests in that no explicit good quality reference is introduced prior to each evaluation. The experimental design followed to run this type of tests should be the n x n graeco-latin square, suitable to take into account four variables with the same degree of freedom (n-1): test condition, order of presentation, conversation task and subject. It is necessary to stress the point that the choice of an experimental design and of the subsequent model for variance analysis determine completely the effects which could be estimated from a statistical point of view. In its normalised version for purpose of overall quality evaluation, non-experienced subjects are used for these tests. It is possible to conduct simplified conversation tests with different applications: diagnostic or validation of the device under test. In this specific case, professional subjects (operators of telephonometry, experts,..) give their opinion on several criteria; no graeco-latin square design is requested, only one conversation task is used. These simplified tests are not normalised and not statistically validated (4 or 5 collected votes only). |
d6e73de470353c70c32b5dcc77258e0b | 101 110 | 4.3.4.1.3 Proposed test method | The proposed methodology adapted from the Ultimate Test Set is not normalised and even not yet validated. This set of tests should be considered as diagnostic tests such as those run during the development phase of a coder which are not documented in ITU-T. The procedure uses five point category-judgement scales which should be conformed to recommended existing ones: quality, listening effort (and derived talking effort) impairment scale, difficulty scale, loudness preference, etc. The fact that no experimental design, and thus, no control of experimental variables are given, along with the small number of collected votes do not allow a statistical justification of the results. ETSI ETSI TR 101 110 V8.0.0 (2000-04) 40 (GSM 03.58 version 8.0.0 Release 1999) It should be noted that, within ITU-T SG 12, a new recommendation which requires similar methodological clarification, is under study in Question 14/12. The future recommendation (P.SEEC Speech Evaluation of Echo Canceller) is devoted to the evaluation of the quality performance of echo cancellers and will form a basis of work on other active signal processing devices as those implemented in handsfree terminals. A provisional list of subjective possible procedures along with their applications has been already elaborated which try to prioritize the different methodologies, e.g. Conversation Tests, Talking &and Listening Tests, Third Party Listening Tests, depending of their specific goals, e.g. Overall Opinion and/or Quality, Diagnostic, Parameter Value Selection. It is clearly said in the reports of ITU-T SG12 Question 14/12 on this matter that the unique reference will always be the "true, classic" conversation test, as the only one test and that the relationship between Conversation MOS scores and other results should be understood and firmly stated. |
d6e73de470353c70c32b5dcc77258e0b | 101 110 | 4.3.4.2 Subjective tests extracted from "Ultimate test set" | The following subclause presents a summary of Race R2072 WP2.2 contribution titled "The Ultimate Test Set". This text defines several tests: - Objective tests; - Subjective tests: - Talker test comprises tests that should reveal any degradations that will hinder the talker in speaking; - Listener test aims at revealing any degradations that will hinder the listener in listening. Part 1 is described in the following (speech and silence; CVC). Part 2:see ITU-T Recommendation. - Interactive test (interruption and interaction) is designed to assess the communication link when talker and listener are involved in a highly interactive test, where the respective roles of talker and listener are frequently swapping; - Conversation test should reveal the subjective quality of the link when it is used in a normal conversation. (See ITU-T Recommendation P.800). The subjective tests, except those defined in the ITU-T recommendations are described in annex A. 1) Talker test Table 10: Talker test Test Type Parameters tested Free monologue Noise Counting test Echo The test can be performed by experimented subjects. It is recommended that at least four different talkers (two male, 2 female). The listener can be replaced by a HATS. 2) Listener test Table 11: Listener test Test type Ambient noise Parameters tested Speech and silence Echo Speech and silence distant ambient noise Influence of the distant ambient noise Logatoms (CVC) DAV ETSI ETSI TR 101 110 V8.0.0 (2000-04) 41 (GSM 03.58 version 8.0.0 Release 1999) The test can be performed by experimented subjects. It is recommended that at least four different talkers (two male, 2 female). The talker can be replaced by a HATS placed in the standard conditions. 3) Interaction and interruption tests The test is made between two partners. It is mainly performed to test the behaviour of speech processor included in the handsfree terminal. Both subjects have active parts, they are faced to two simultaneous tasks, ex: counting and listening. This is difficult task which cannot be asked to naive subjects. Moreover, the evaluation criteria related to voice suppression, voice clipping, etc, is out of the possible interpretation of non experienced subjects. According to the classification achieved in ITU-T SG 12 Question 14/12 , this test should remain a diagnostic tool performed by experts. ETSI ETSI TR 101 110 V8.0.0 (2000-04) 42 (GSM 03.58 version 8.0.0 Release 1999) Annex A: Subjective Tests from UTS TALKER TEST / SLOW COUNTING The purpose of the test is to reveal degradations that hamper the talker in speaking. The talker is asked to slowly count out from 1 to 20 and to answer to the following questionnaire: Q 1. Did you hear any echo ? Score Echo annoyance 5 degradation is inaudible 4 degradation is audible but not annoying 3 degradation is slightly annoying 2 degradation is annoying 1 degradation is very annoying Q 2. How would you rate the quality of the connection from the point of view of a talker? Score Quality Scale 5 excellent 4 good 3 fair 2 poor 1 bad TALKER TEST / FREE MONOLOGUE It is asked to have a free monologue. You are completely free in choosing whatever you find meaningful to test the quality. For example, you could whisper, make an exclamation or just talk n a normal manner. While talking, you should listen for any degradations causing difficulty in talking. Q 3. What is your opinion on the amount of effort required to talk caused by the presence of any noise: hum, clicks, beeps or tones? Score Talking effort scale 5 complete relaxation possible, no effort required 4 attention necessary, no appreciable effort required 3 moderate effort required 2 considerable effort required 1 extreme effort: talking impossible Q 4. How would you rate the quality of the connection from the point of view of a talker? Score Quality scale 5 excellent 4 good 3 fair 2 poor 1 bad ETSI ETSI TR 101 110 V8.0.0 (2000-04) 43 (GSM 03.58 version 8.0.0 Release 1999) LISTENER TEST / SPEECH AND SILENCE TEST You are going to hear 4 samples of speech and silence. After listening to each sample, answer to the questionnaires. Q 5. Did you hear any echo? Score Sample 1 Sample 2 Sample 3 Sample 4 Impairment scale:Echo annoyance 5 5 5 5 degradation is inaudible 4 4 4 4 degradation is audible but not annoying 3 3 3 3 degradation is slightly annoying 2 2 2 2 degradation is annoying 1 1 1 1 degradation is very annoying Q 6. How would you rate the quality of the connection from the point of view of a listener? Score Sample 1 Sample 2 Sample 3 Sample 4 Quality Scale 5 5 5 5 excellent 4 4 4 4 good 3 3 3 3 fair 2 2 2 2 poor 1 1 1 1 bad Q 7. What is your opinion on the amount of effort required to talk caused by the presence of any noise: hum, clicks, beeps or tones? Score Sample 1 Sample 2 Sample 3 Sample 4 Listening effort scale 5 5 5 5 complete relaxation possible, no effort required 4 4 4 4 attention necessary, no appreciable effort required 3 3 3 3 moderate effort required 2 2 2 2 considerable effort required 1 1 1 1 extreme effort Ambient noise is generated in the distant room Q 8. How do you perceive the distant ambient noise? Score Sample 1 Sample 2 Sample 3 Sample 4 Impairment scale: distant ambient noise annoyance 5 5 5 5 degradation is inaudible 4 4 4 4 degradation is audible but not annoying 3 3 3 3 degradation is slightly annoying 2 2 2 2 degradation is annoying 1 1 1 1 degradation is very annoying ETSI ETSI TR 101 110 V8.0.0 (2000-04) 44 (GSM 03.58 version 8.0.0 Release 1999) LISTENER TEST / CVC The test is designed to test the Voice Activity Detector (VAD) Q 9. What is your opinion on the amount of listening effort? Score Listening effort scale 5 complete relaxation possible, no effort required 4 attention necessary, no appreciable effort required 3 moderate effort required 2 considerable effort required 1 extreme effort INTERUPTION TEST At the other side of the telephone connection, a talker is continuously generating speech. You are asked to interrupt the talker by uttering sounds like "uuhu", "yes", "mmm",...to detect degradations caused by the equipment. Q 10. How do you rate the amount of voice suppression ? Score Voice suppression perception 5 no voice suppression perceived 4 voice suppression is audible but not annoying 3 voice suppression is slightly annoying 2 voice suppression is annoying 1 voice suppression is very annoying Q 11. How would you rate the overall communication quality ? Note Quality scale 5 excellent 4 good 3 fair 2 poor 1 bad ETSI ETSI TR 101 110 V8.0.0 (2000-04) 45 (GSM 03.58 version 8.0.0 Release 1999) INTERACTION TEST/ COUNTING TEST The two partners are at each side of the link and are asked to take turns uttering the numbers from 1 to 20 (one uttering the even, the other the odd numbers) as quickly as possible. The counting can be repeated as often as the partners consider necessary. While counting they should listen for degradations caused by voice clipping and delay. Q 12. How would you rate the amount of voice clipping? Score Amount of voice clipping 5 no voice clipping perceived 4 voice clipping is audible but does not influence the task 3 voice clipping slightly causes difficulty to perform the task 2 voice clipping causes considerable effort to perform the task 1 voice clipping causes extreme effort: no conversation possible Q 13. Do you perceive any delay? Score Amount of delay 5 no delay perceived 4 delay perceived but does not influence the task 3 delay slightly causes difficulty to perform the task 2 delay causes considerable effort to perform the task 1 delay causes extreme effort: no conversation possible Q 14. How would you rate the overall communication quality? Score Quality scale 5 excellent 4 good 3 fair 2 poor 1 bad ETSI ETSI TR 101 110 V8.0.0 (2000-04) 46 (GSM 03.58 version 8.0.0 Release 1999) Annex B: Bibliography B.1 References of TD presented in SMG2/ad hoc 03.50 meetings 30/95. FT/CNET/ Monfort:Analysis of information available on handsfree in mobile (vehicle) environment. 9/94. Telia/ Karlsson: Measurement on a GSM handsfree telephone). 58/95. BT/ Goetz: Handsfree test environment. 42/95 FT/CNET/ Monfort: Test to be applied to ETS 300 540. 49/95. BT/ Goetz: Advisory text for installation of handsfree MS in a vehicle environment. 60/95. Head Acoustics/ Gierlich: Cost estimation for a vehicle simulator for handsfree testing. 12/95. Matra Communications/ J. Boudy: Additional speech processing delay for the use of Handsfree. 23/95. Matra Communications/ J. Boudy: Delay figures for HF processing. 35/95. Matra Communications/ J. Boudy: Additional delay for HF processing. 77/95. Ericsson: Extra delay for handsfree signal processing. 41/95 Matra Communications/ J. Boudy: Evaluation methodology for full duplex acoustic echo controllers developed within Freetel Project. 11/94. BT/I.Goetz: Assessing GSM speech quality. B.2 References from subclause 4.1 [1] Background acoustic noise reduction in mobile telephony. R.A. Goubran, H.M. Hafez. 36th IEEE Vehicle Technology Conference. [2] Speech enhancement for mobile telephony. M.M. Goulding, J. Bird. IEEE Transactions on vehicular technology, Vol 39, N°4, November 1990. p 316-326. [3] Contribution à l'amélioration des performances d'un radiotelephone mains-libres à commande vocale. C. Baillargeat. Thésis (18 April 1991). [4] Acoustic noise analysis and speech enhancement for mobile radio applications. N. Dal Degan, C. Prati. Signal Processing 15, N°1, July 1988, p43-56. [5] On the influence of front end processing schemes on the GSM codec behaviour in the context of handsfree radiotelephony. P. Scalart, A. Benamar. [6] Noise spectra CCITT blue book, 1988, Volume 5. Supplement n°13. [7] Binaural Measurements of loudness as a parameter in the evaluation of sound quality in automobiles. Greg Michel, Gordon Ebbitt, Bruel and Kjaer Instruments Inc. [8] Hands-free and handset mobile telephoning: simulated in situ assessment of telephonic signals and noise using HATS. Bruel & Kjaer. CCITT COM XII-51. July 1990. [9] Subjective evaluation of quality in car hands-free radiotelephone situation. France Télécom/CNET. CCITT SG XII.,D.95, Brasilia, September 1991. ETSI ETSI TR 101 110 V8.0.0 (2000-04) 47 (GSM 03.58 version 8.0.0 Release 1999) [10] Subjective evaluation of quality of communications in car hands-free radiotelephone context. UIT-T Com 12-12, March 1993. France. B.3 References from subclause 4.2.1 [1] State of art in Acoustic Echo Cancellation .A. Gilloire, E. Moulines, D.Slosk, P. Duhamel. Published in "Digital signal processing in telecommunications". B.4 References from subclause 4.3.1 "Sound Fields Around the Head/Handset System And Their Exploitation For Pressure Gradient Noise Cancellation": M.P. Hollier, Institute Of Acoustics Proceedings Vol. 12, Part 10 (1990). "Communications in noise - a performance ranking metric", Hollier M P, Guard D R, Hawksford M O, BT Technology Journal, Vol. 10, N° 4, October 1992. 'Outline Proposal For A Noise Rejection Specification For 1/2-Rate GSM' M. Hollier, BTL, Tdoc SMG-02 03.50 32/95. "Echo Loss Of GSM Mobiles"; I. Goetz, SMG-02 03.50 Tdoc 4/93. "Speech & Speech Model Signals: A Comparison" ITU-TS SGXII Q7/12 & 13/12 Contribution 14, March 1993. B.5 References from subclause 4.3.3 [1] "The Ultimate Test Set", RACE R2072 WP2.2 Contribution, PTT Research, Holland. [2] "Status of the Standardization activities on Acoustic Echo Control", A. Gilloire, Signal Processing, nø27, pp.273-279, 1992. [3] "Performance Evaluation of Acoustic Echo Control: Required Values and Measurements Procedures", A. Gilloire, 3rd Int. Workshop on Acoustic Echo Control, Sep. 93, Lannion-France. [4] "EVALUATION METHODOLOGY FOR FULL-DUPLEX ACOUSTIC ECHO CONTROLLERS" Contribution ITU-T COM 12-40 (September 1994)), Source: FREETEL Project. [5] "FREETEL FINAL REPORT ", ESPRIT IV Project n° 6166. ETSI ETSI TR 101 110 V8.0.0 (2000-04) 48 (GSM 03.58 version 8.0.0 Release 1999) Annex C: Document change history SPEC SMG# CR PHA SE VERS NEW_VE RS SUBJECT 03.58 s23 new R97 2.0.0 5.0.0 Handsfree test methods 03.58 s26 R97 5.0.0 6.0.0 R97 version change 03.58 s26 R97 6.0.0 6.0.1 Editorial changes for publication 03.58 s29 R98 7.0.0 Specification version upgrade to Release 1998 version 7.0.0 03.58 S31 R99 8.0.0 Specification version upgrade to Release 1999 version 8.0.0 ETSI ETSI TR 101 110 V8.0.0 (2000-04) 49 (GSM 03.58 version 8.0.0 Release 1999) History Document history V8.0.0 April 2000 Publication |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 1 Scope | |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 1.1 Background | CE Mandate M/239 [8] was accepted in principle by TA 25 in October 1996 and the Terms of Reference of Special Task Force (STF) 109 [9] were agreed by Board Number 6 (B6(97)08). Under the mandate CEN/CENELEC/ETSI have been asked to draw up a programme of standards to complement Eurocontrol's programme of technical specifications. The present document has been produced by a STF that was set up to deal with one of the items given in an annex to the mandate. The purpose of the STF was to carry out a study on the feasibility of standardizing Self-organizing Time Division Multiple Access (STDMA) mode 4 and to identify possible requirements for future activity. The overall strategy is to develop a standard in Europe for STDMA to be installed on aircraft, initially on a voluntary basis. The ETSI standard would be derived from and, wherever possible, maintain compatibility with, the draft Standards and Recommended Practices (SARPs) being developed by the International Civil Aviation Organization (ICAO) for VHF Digital Link (VDL) mode 4 (the ICAO term for STDMA). These ICAO standards will be referred to throughout this document as "VDL mode 4 draft SARPs". It should be noted that the European Commission is funding the STF work (100 %) because it wishes to speed up the standardization process for STDMA. The Commission is anxious to determine if the standard can be progressed and the predicted timescales for it to become an ETSI European Norm (EN). The aim of the work was to analyse the available technical specification and produce recommendations for what actions ETSI should take to: • transfer the technical specification into a standard; • complete all the required interfaces; • identify areas needing particular attention; • define the appropriate time schedule for any future work. The present document reports the results of the work. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 1.2 Organization of the report | The present document is organized as follows: • Clause 2 lists the references. • Clause 3 lists definitions, symbols and abbreviations. • Clause 4 provides a summary of STDMA, describing the system requirements, the main features of the system and the system concept. • Clause 5 describes the approach that was taken by the STF in the review of STDMA. • Clause 6 presents the results of the review. It contains the following subclauses: • a summary of the perceived need for an STDMA standard; • a summary of procedural issues that will have to be addressed to achieve standardization, including the relation to other standards activities and the possible ETSI process that could be used to produce standards material; • a summary of the standardization issues that will have to be addressed to improve the presentation of the existing ICAO standard in order to bring it into line with ETSI best practice; TR 101 130 V1.1.1 (1997-11) 8 • a summary of the technical issues that will have to be addressed to complete the technical specification of the existing system. • Clause 7 presents a proposed work plan for STDMA standardization. • Clause 8 presents the conclusions and recommendations of the study. More detailed information on the design principles of STDMA, the procedural issues, the standardization issues and the technical issues are presented in annex A, B, C and D respectively. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 2 References | References may be made to: a) specific versions of publications (identified by date of publication, edition number, version number, etc.), in which case, subsequent revisions to the referenced document do not apply; or b) all versions up to and including the identified version (identified by "up to and including" before the version identity); or c) all versions subsequent to and including the identified version (identified by "onwards" following the version identity); or d) publications without mention of a specific version, in which case the latest version applies. A non-specific reference to an ETS shall also be taken to refer to later versions published as an EN with the same number. [1] "VDL mode 4 Standards and Recommended Practices - DRAFT", Version 5.4, 21 March 1997 for presentation to Aeronautical Mobile Communication Panel (AMCP), Working Group D, Seventh Meeting, Madrid, Spain, 8 - 17 April 1997. [2] "VDL mode 4 Manual - DRAFT", Version 1.0, 21 March 1997 for presentation to Aeronautical Mobile Communication Panel (AMCP), Working Group D, Seventh Meeting, Madrid, Spain, 8 - 17 April 1997. [3] "Changes to VDL mode 4 SARPs in version. 5.4 with respect to version 4.0", Information paper, 21 March 1997 for presentation to Aeronautical Mobile Communication Panel (AMCP), Working Group D, Seventh Meeting, Madrid, Spain, 8 - 17 April 1997. [4] "A proposal for ATN over VDL mode 4", Information paper, 21 March 1997 for presentation to Aeronautical Mobile Communication Panel (AMCP), Working Group D, Seventh Meeting, Madrid, Spain, 8 - 17 April 1997. [5] "The choice of modulation scheme for VHF NABS: A comparison of GFSK and D8PSK", paper presented by H Westermark to WG-D of GNSSP 4th meeting, Australia 17 - 28 February 1997. [6] "Spectrum requirements for a VHF navigation augmentation broadcast system (NABS) for D8PSK and GFSK modulation schemes", paper presented by H Westermark to WG-D of GNSSP 4th meeting, Australia 17 - 28 February 1997. [7] "Flight trials of GFSK 19.2kbps radios", paper presented by H Westermark to WG-D of GNSSP 5th meeting, Montreal 26 May - 6 June 1997. [8] Commission mandate M/239: "Mandate to CEN/CENELEC/ETSI for standardization, and a study, in the field of Air Traffic Management Equipment and Systems", 5 September 1996. [9] "Terms of reference for STF ER on Self organizing time division multiple access system (STDMA) for Aeronautical VHF communications" as agreed by Board 6 (B6(97)08). TR 101 130 V1.1.1 (1997-11) 9 [10] "Enhanced TDMA for a VHF Datalink System matching the future European Air Traffic Management System requirements (E-TDMA): Report on System Requirements (WP1)" CEC DGXIII TREATY 8 report, 28 April 1997. [11] "Maintaining the Time in a Distributed System", Keith Marzullo. 2nd ACM Symposium on Principles of Distributed Computing. Montreal, August 1983 pp 295-305. [12] "Fault-Tolerance Clock Synchronization", Joseph Halpern, Barbara Simons, Ray Strong. 3rd ACM Symposium on Principles of Distributed Computing Systems. Vancouver, Canada, August 1984, pp 68-74. [13] "Byzantine Clock Synchronization", Leslie Lamport, P. M. Melliar-Smith, 3rd ACM Symposium on Principles of Distributed Computing Systems. Vancouver, Canada, August 1984, pp 68-74. [14] "Understanding protocols for Byzantine clock Synchronization", F Schneider. Technical Report 87-859, Dep of Computer Science, Cornell University, August 1987. [15] "Integrating External and Internal Clock Synchronization", Christof Fetzer, Falviu Christian. 15th International Conference on Distributed Systems, Vancouver, Canada May 1995. [16] "An optimal Internal Clock Synchronization Algorithm", Christof Fetzer and Flaviu Christian. 10th Annual IEEE Conference on Computer Assurance, Gaitersburg, MD, June, 1995. [17] "Lower Bounds for Function Based Clock Synchronization", Christof Fetzer and Flaviu Christian. 14th ACM Symposium on Principles of Distributed Computing, Ottowa, CA, August 1995. [18] "Minutes of Airline Workshop, 18-19 December 1996 in Saltsjöbaden, Stockholm", Information paper, 21 March 1997 presentation to Aeronautical Mobile Communication Panel (AMCP), Working Group D, Seventh Meeting, Madrid, Spain, 8 - 17 April 1997. [19] ISO 8208 (1995): " Information technology - Data communications - X.25 Packet Layer Protocol for Data Terminal Equipment". |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 3 Symbols and abbreviations | |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 3.1 Symbols | For the purposes of the present document, the following symbols apply: dBm dB relative to 1 mW kbps kilo bits per second. (unit of transmission rate) kHz kilo Hertz (frequency unit) nmi nautical mile. (distance unit equal to 1 832 metres) |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 3.2 Abbreviations | For the purposes of the present document, the following abbreviations apply: ACAS Airborne Collision Avoidance System ADS-B Automatic Dependent Surveillance Broadcast ADS-C Automatic Dependent Surveillance Contract ADSP Automatic Dependant Surveillance Panel AOC Airline Operators Communications ASAS Airborne Separation Assurance ATC Air Traffic Control ATM Air Traffic Management TR 101 130 V1.1.1 (1997-11) 10 ATN Aeronautical Telecommunication Network AWOP All Weather Operation Panel CCI Co-Channel Interference CDTI Cockpit Display of Traffic Information CFP Cellular Frequency Planning CPDLC Controller pilot data link communication CNS Communication, Navigation and Surveillance D8PSK Differentially Encoded 8 Phase Shift Keying DLS Data Link Service DME Distance Measuring Equipment DoS Directory of Service FIS Flight Information Service GA General Aviation GFSK Gaussian Filtered Frequency Shift Keying GNSS Global Navigation Satellite System GNSSP Global Navigation Satellites Systems Panel GSC Global Signalling Channel HF High Frequency LLC Logical Link Control LME Link Management Entity MAC Media Access Control MASPS Minimum Aviation System Performance Standards MOPS Minimum Operational Performance Specification NEAN North European ADS-B Network NEAP North European CNS/ATM Applications Project RFP Reuse Frequency Planning RSSI Received Signal Strength Indication SARPs Standards and Recommended Practices SICASP Secondary Surveillance Radar Improvements and Collision Avoidance Systems Panel SMGCS Surface Movement Guidance and Control System SNR Signal to Noise Ratio STDMA Self-organizing Time Division Multiple Access STF Special Task Force T The baud period or 1/baud rate TC Technical Committee TIS Traffic Information Service TDMA Time Division Multiple Access UTC Universal Co-ordinated Time VDL VHF Digital Link VHF Very High Frequency VSS VDL mode 4 Specific Services |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 4 Summary of STDMA | |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 4.1 Introduction | This subclause sets provides a summary of STDMA, describing the system requirements, the main features of the system and the system concept. This subclause is based on the material provided in references [1], [2], [3] and [10]. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 4.2 System requirements | STDMA is designed to meet a system requirement that is to provide a data link technology that enables a range of Air Traffic Management (ATM) applications, including: • Automatic Dependent Surveillance-Broadcast (ADS-B): In ADS-B an aircraft broadcasts its position and other related data, such as ground speed and track, to all other mobile and ground based users in the vicinity. ADS-B potentially enables many new user applications including Cockpit Display of Traffic Information (CDTI), TR 101 130 V1.1.1 (1997-11) 11 station keeping (i.e. one aircraft following another at a certain distance), augmented Air Traffic Control (ATC) surveillance and airborne separation maintenance. ADS-B is the primary application of STDMA. Note that there is another type of ADS system, known as ADS-contract (ADS-C), in which position reports are set up and transmitted using two-way point-to-point communication links. • Differential Global Navigation Satellite System (GNSS) augmentation: When using GNSS data for navigation or surveillance, a GNSS augmentation system may be used to ensure the quality of the position data. An uplink broadcast data link is one way to provide GNSS augmentation signals, which provide information on the quality of the GNSS signals and correction data to overcome errors and inaccuracies in the signals from the satellites. • Surface Movement Guidance and Control System (SMGCS): SMGCSs provide surveillance of ground traffic at airports. The traffic may include ground vehicles and taxiing or parked aircraft. The application requires the exchange of surveillance and other types of data between all users in the vicinity of the airport. SMGCS is essentially a ground-based application of ADS-B. • Controller pilot data link communication (CPDLC): CPDLC provides pilot-controller digital communications for future applications. CPDLC requires a two way data link system for its operation. Such a data link could also support other point to point applications such as Airline Operators Communications (AOC) providing information exchange between aircraft and airlines. CPDLC is an application of the Aeronautical Telecommunications Network (ATN). The VDL mode 4 draft SARPs define functions that could support ATN- compliant point-to-point communication and hence, if these functions are incorporated into the ETSI standard, STDMA could become a mobile subnetwork of the ATN which allows it to support CPDLC and other ATN applications. • Uplink broadcast information: STDMA can be used to provided uplinked information on, for example, meteorological data, Flight Information Services (FIS), Traffic Information Services (TIS) and other broadcast information. Broadcast applications are provided outside of the framework of the ATN. Wherever possible, these applications should be supported in a variety of airspace conditions such as busy continental regions and low density oceanic airspace. The exact system requirements for a data link system are difficult to define for a variety of reasons including: • Many of the application requirements are not well defined and the subject of ongoing debate. • The mix of data link technologies required to provide a "system solution" to the application requirements, taking account of such factors as: • safety and certification requirements (which may lead to the requirement to distribute different applications over a number of physically independent data links); • cost (which may lead to the requirement to reduce the number of data link technologies for which aircraft must be equipped), is not determined and could probably only be finalized once the application requirements are finalized. Hence, system requirements can currently only be specified against example operational scenarios. For the purposes of discussing the requirements for STDMA, it will be assumed that STDMA will support an "ADS-B rich" scenario in which repetitive broadcast of position information is the dominant data link load. The DGXIII TREATY 8 report [10] derives an example scenario, which also includes applications which give rise to a small point-to-point data load. Typical message transfer requirements based on the DGXIII TREATY 8 report [10] for an en-route scenario are contained in table 1. TR 101 130 V1.1.1 (1997-11) 12 Table 1: Typical message transfer requirements for an en-route traffic scenario Message length Quality of Service Priority Message Frequency (messages per hour) Message type Short Low Routine 456 Point-to-point Medium Routine 912 Point-to-point High Critical Routine 5 472 3 648 Point-to-point Point-to-point Very high Critical 194 940 Broadcast Medium Low Routine 3 648 Point-to-point High Critical Routine 456 456 Point-to-point Point-to-point Very high Critical 10 620 Broadcast Long Low Routine 2 280 Point-to-point High Routine 2 736 Point-to-point Very long High Routine 456 Point-to-point In table 1, the terms have the following meanings: • Message length: • short: less than 20 octets; • medium: between 20 and 200 octets; • long: between 200 and 3 000 octets; • very long: over 3 000 octets. • Quality of service: • low: message delivery time greater than 20s; • medium: message delivery time between 10 and 20s; • high: message delivery time between 5 and 10s; • very high: message delivery time less than 5s. • Priority: • Critical: relates to emergencies and flight safety; • Routine: other essential messages. The data in the table assumes that there are 570 aircraft in an area defined by a circle of radius 160 nautical miles (typical of traffic densities predicted for core European airspace in 2005). There is great potential for different system requirements than those set out above. In particular: • the communications load requirements will fall substantially over oceanic and non-core continental airspace because the traffic density is lower; • it would be possible to envisage a scenario that has a low ADS-B requirement but which provides a high level of point-to-point communications. If possible, the data link solution should be flexible enough to meet these extremes. TR 101 130 V1.1.1 (1997-11) 13 |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 4.3 Candidate data link technologies | There are a number of data link technologies that could individually, or in combination, meet some or all of the requirements outlined in subclause 4.2. These include: • VDL mode 2, which provides a ground/air Very High Frequency (VHF) link using CSMA protocols. It has been standardized by ICAO, is targeted at applications for AOC and is not expected to be suitable for safety critical communications. VDL mode 2 provides ATN data communications only. • VDL mode 3, which has been developed as a possible extension of the VDL to provide Time Division Multiple Access (TDMA) channel access for ground/air ATN data and voice communication. • VDL mode 4 (the ICAO term for STDMA), which has been developed to provide support to ADS-B and to also support data communications. VDL mode 4 provides ATN and non-ATN data communications. • Mode S squitter. This L-band system was originally developed to support Airborne Collision Avoidance System (ACAS) applications but is also has the potential to support ADS-B applications. STDMA is the subject of the present document. However, standardization of STDMA needs to be justified against a background of technologies that might provide alternative routes for meeting the system requirements set out in subclause 4.2. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 4.4 Key features of STDMA | STDMA is a mode for Communication, Navigation and Surveillance (CNS) systems using either a Differentially Encoded 8 Phase Shift Keying (D8PSK) or a Gaussian Filtered Frequency Shift Keying (GFSK) modulation scheme on standard 25 kHz VHF radio channels and a Self-organizing Time Division Multiple Access (STDMA) scheme . The STDMA technology was invented in Sweden. This data link has been designed to support repetitive short air-to-air position report broadcasts (ADS-B) as well as to support long and non-repetitive transmissions and ATN services. STDMA has the potential to cover a wide variety of data exchanges applications (ADS-B, Differential GNSS, ATN, CPDLC, FIS-broadcast, AOC, CDTI, etc.). Figure 1 illustrates the services and applications that could be provided by STDMA (based on functions included in VDL mode 4 draft SARPs [1]): STDMA Services ATN services VDL Mode 4 specific Core functions supported by VDL Mode 4 User applications ADS-B GNSS-P ATN Data broadcast ground-air air-air Position broadcast ground-air air-air End-to-end communications ground-air Cockpit Display of Traffic Information (CDTI) Separation assurance Precision approach monitor Surface movement surveillance Secondary navigation Flight Information Service broadcast (FIS-B) Traffic Information Service - broadcast (TIS-B) Precision navigation Air-air communications Air-air trajectory negociation Non-ATN ground-air communications Controller Pilot Data Link Communication (CPDLC) ADS contract (ADS-C) End-to-end communications ground-air air-air Figure 1: STDMA communications services and example applications TR 101 130 V1.1.1 (1997-11) 14 The main concepts of STDMA are: • A TDMA media access system using a large number of short time-slots. • Distributed time synchronization: Universal Co-ordinated Time (UTC) time (provided by GNSS receivers or other means) is used to synchronize to the time-slots. • Managed access to the time-slots: each user maintains information on the planned usage of all timeslots. This information is initially gathered by listening to the channels before attempting to access the data link, and thereafter it is constantly updated. The information is used to decide the time-slots in which a station will transmit data. The intention to use one or more slots for data transmissions is announced using slot reservation protocols. Decisions to transmit data may be made by a mobile user operating autonomously or under the direction of a ground station. • Adaptive slot selection mechanism: for some types of applications, each user applies an algorithm to avoid long- term slot collisions (i.e. the same slot selected for transmissions by several users). This enables applications such as ADS-B to operate without the presence of a ground infrastructure. • Position reports broadcast: each user regularly transmits its identity and position (position may be provided by a GNSS receiver or other means) in a synchronization burst. This information is required for communications management, e.g. to provide connectivity information for air-to-air communications. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 4.5 STDMA system concept | Each STDMA user may tune to any of the 25 kHz frequency channels from 108 to 136,975 MHz for receiving data, and from 112 to 136,975 MHz for transmitting or receiving data. The band from 108 to 112 MHz is reserved for ground transmissions only. The band from 112 to 136,975 MHz may be used for ground or airborne transmissions. It is envisaged that two 25 kHz channels, to be known as Global Signalling Channels (GSC), will be assigned a priori to STDMA. These channels will be used by ground stations for system management, by transmitting Directory of Service (DoS) messages. DoS messages announce the availability of services on different channels. The channels may also be used to support some uplink broadcasts applications and ADS-B reporting by some aircraft. The use of two channels (as opposed to one) is proposed to allow continued operation in the case of unintentional jamming or blocking of one of the channels. One GSC channel has been proposed as 136,95 MHz, but the other is not defined. The following airborne architectures of STDMA have been proposed: • Commercial air transport: Aircraft has capability to receive on three channels and transmit on one. Two of the channels are assigned to the GSCs and the third to data communications. • General aviation aircraft: Aircraft has capability to receive on two channels and transmit on one. The two channels are assigned to the GSCs and the aircraft does not have a data capacity. These architectures define minimum levels of functionality for different users with different requirements. A higher level of functionality may be present on an aircraft and, for example, figure 2 shows a possible architecture that supports reception on four channels and transmission on two. Rx Rx Tx Equipment 1 Rx Rx Tx Equipment 2 Figure 2: Possible airborne architecture for commercial aircraft Additional channels may be employed specifically for a particular application, or group of applications. A number of options are possible as illustrated in table 2. These options may be applied individually, or in combination, recognizing the limited number of receivers available on aircraft. TR 101 130 V1.1.1 (1997-11) 15 Table 2: Possible options for use of additional channels Description Local channel(s) for ADS-B reporting Channel for wide area en-route end-to-end communications Channel for local terminal end-to-end communications Channel for local uplink broadcast applications, including TIS-B, GNSS augmentation and FIS-B Channel for SMGCS Physical layer characteristics are assumed to be: • GSCs are expected to use a GFSK 19,2 kbps modulation scheme. • Data channels will use a GFSK 19,2 kbps or differentially encoded 8 phase shift keying (D8PSK) 31,5 kbps modulation scheme. Each user equipment will need to comprise at least (minimum hardware configuration) one transmitter which may tune alternatively to the used frequencies (GSCs and dedicated channels) and two receivers which monitor each GSC full- time. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 5 Approach to the review of STDMA | This clause describes the approach that was taken by the STF in the review of STDMA. An STF of three experts was set up to carry out the required work. The study team comprised expertise in STDMA, air traffic management issues, communication systems and previous experience of the relevant ETSI processes. The study took place between 12 August 1997 and 29 September 1997. The following phased approach was used: • In phase 1, the available STDMA documentation was reviewed and a list of issues generated for further study. • In phase 2, further work was carried out on each issue. This phase included consideration of a possible ETSI standardization process and the generation of a recommended work plan. • In phase 3, the results of the work were collated and summarized in a draft technical report. The list of documents analysed are references [1], [2], [3], [4], [5], [6]. Note that the purpose of the study was not to provide solutions for issues raised in the analysis. Rather it was intended to identify the key issues and to produce a study report that could be used to decide whether to proceed with STDMA standardization and, if so, to define the scope of work necessary to complete the standard. A draft TR was presented for consideration at the Technical Committee (TC) ERM meeting on 6th to 10th October 1997. Comments from that meeting were then incorporated to produce a final TR. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6 Results of the review of STDMA | |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.1 Introduction | This clause presents the results of the review of STDMA. It contains the following subclauses: • a summary of the perceived need for an STDMA standard; • a summary of procedural issues that will have to be addressed to achieve standardization, including the relation to other standards activities and the possible ETSI process that could be used to produce standards material; • a summary of the standardization issues that will have to be addressed to improve the presentation of the existing ICAO standard in order to bring it into line with ETSI best practice; TR 101 130 V1.1.1 (1997-11) 16 • a summary of the technical issues that will have to be addressed to complete the specification of the existing system. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.2 The need for STDMA | |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.2.1 Market need | Some airlines have expressed a need for the applications supported by STDMA (see "Minutes of Airline Workshop" [18]), establishing a market need for the system. In addition, the system has the potential to offer a low cost data link supporting a range of applications that is affordable by General Aviation users. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.2.2 Technical need | Annex A summarizes the results of an analysis of the design principles of STDMA. The following characteristics of STDMA have been identified during the review by inspection of the draft VDL mode 4 SARPS [1]: • both broadcast and point-to-point functions; • a concept of operation which uses Reuse Frequency Planning (RFP) in which users share access to common channels, a concept which seems particularly suited to ADS-B and other broadcast functions; • support for Cellular Frequency Planning (CFP), which may be more appropriate for point-to-point functions; • distributed synchronization, which allows the VDL to work without ground infrastructure; • ability to also support centralized synchronization and access control; • short time-slots well suited for repetitive broadcast applications (longer messages can be sent by using a block of slots); • a flexible channel access scheme that offers enhanced efficiency compared with random access schemes through the use of reservation protocols; • world-wide signalling channels combining traffic control and data exchanges for various applications; • spectral efficiency in which two alternative modulations schemes, D8PSK and GFSK, are supported. The analysis has shown that STDMA appears well suited for repetitive short messages broadcast (position reports) as this function is an inherent core part of the data link. A decision to standardize STDMA must be taken against a background of other data link technologies that could provide a better solution to the requirements of future applications. As was described in subclause 4.1, there are a wide range of future communications, navigation and surveillance ATM applications, which can provide benefits to users in oceanic, core continental and low density continental airspace. Realization of these applications will require the development of enabling communications technology or combinations of technology. The detailed requirements for these applications are unknown and subject to change. However, it is likely that the requirements for the enabling data link technology will be diverse and may include: • operating with and without ground infrastructure; • providing ground/air and air/air communication; • providing point/point and broadcast communication functions; • supporting safety critical communication; • using a robust handover mechanism between coverage cells, particularly for safety critical applications; • maximizing spectrum efficiency. TR 101 130 V1.1.1 (1997-11) 17 The capability of VDL modes 2, 3 and 4 and mode S squitter to support the applications listed in subclause 4.2 is a subject of debate. STDMA has certain features that are not provided by any other proposed VHF data link in ICAO. These unique functions include: • operation with no ground infrastructure; • air-to-air communications, both with and without the presence of ground stations; • broadcast communications, both from ground and airborne users. Note also that this review has not attempted a detailed comparison of how well VDL modes 2, 3 and 4 support end-to- end applications such as AOC, FIS etc. Hence, although STDMA as defined in Draft VDL mode 4 SARPs claims [2] the potential to offer an improved throughput compared with mode 2 because of the use of reservation protocols, there are as yet no model results to support this claim. Similarly, the reviewers have not found any side by side comparisons of mode 3 and mode 4 performance. The reviewers have recommended that any future STDMA standard should be developed in two stages (see subclause 6.3). The first stage would develop the part of the system that would support ADS-B and would enable a side by side comparison with the mode S squitter technology. End-to-end communication would be added in the second stage and it is recommended that a comparison between the VHF modes is made during this stage and, if possible, the mode 4 protocols optimized using experience gained from modes 2 and 3. If possible, opportunities to provide interoperability and transition paths between the different VDL modes should be identified and exploited. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.2.3 Current progress towards standardization | The most recent ICAO activity relevant to VDL standardization was the AMCP Working Group D (WGD) seventh meeting in Madrid from 8 - 17 April 1997. The key results of the meeting were: • Mode 3 VDL standardization: A key service supported by mode 3 is voice communication. Difficulties have been encountered in proving the operation of the vocoder, which is required to provide voice services. The difficulties relate to achieving acceptable discrimination and clarity of voice communication against the high levels of background noise encountered in the cockpit. It was accepted at the meeting that the problems with the vocoder would probably delay the completion of the validation process for mode 3 by at least two years. Note that the development of the data functions of mode 3 have been influenced greatly by the need to provide integrated voice and data. Without the vocoder, it is still possible to standardize VDL mode 3 as an ATN data communications system with higher performance than VDL mode2. • VDL mode 4 (ICAO term for STDMA): The latest version of the SARPs were presented [3]. Discussions were held to decide if mode 4 could enter a validation phase. Objections to the operation of mode 4 were raised on the grounds that the algorithms for frequency sharing were not proven. It should be noted that overall acceptance of mode 4 is hampered by its possible use for ADS-B. Some states, particularly the USA, are committed to using mode S for ADS-B and are unconvinced by the need to consider alternative systems. There is therefore considerable uncertainty as to the way forward for data link standardization within ICAO and this is hampering progress with the development of VDL mode 4: • From a technical viewpoint, it may be better to proceed with standardization in order to provide a stable and validated reference system which can then be considered against evolving operational requirements. Since future requirements will be quite hard to determine in detail for any system considered for standardization, a key feature should therefore be flexibility and growth potential. • On the other hand, an alternative approach is to wait until there is a stable operational requirement before proceeding with standardization activity. This approach would cause significant delay to the progress of VDL mode 4 and greatly reduce its chances of gaining world-wide acceptance. According to current ICAO plans, VDL mode 4 is unlikely to be standardized before 2000. TR 101 130 V1.1.1 (1997-11) 18 |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.2.4 Ongoing trials projects | There are many tests and evaluation projects involving STDMA (note that the equipment used in these trials is based on an early STDMA specification which pre-dates the VDL mode 4 draft SARPs upon which the proposed ETSI standard would be based). Some are finished and some have not yet started, but some of the most relevant ones that are presently ongoing are described below: • NEAN (sponsored by EC DG VII) The largest European STDMA activity is known as the North European ADS-B Network (NEAN). Under NEAN, an ADS-B capability is being created through a network of ground stations and mobile STDMA equipment that is compliant with the emerging VDL mode 4 draft SARPs and which is being installed in commercial aircraft and airport vehicles. The network spans Germany, Denmark and Sweden, and once position reports are received by a ground station they are then distributed throughout the network to air traffic control and other users. There will be 15 ground stations in the NEAN project and 16 aircraft equipped including four 747s, two DC9s, two F28s and a helicopter involved in North Sea operations. Around 30 ground-vehicles will also be equipped. NEAN is a collaborative venture between the German, Danish and Swedish Civil Aviation Administrations and the following aircraft operators: Lufthansa, SAS, OLT, Maersk Helicopters and Golden Air. The UK CAA is leading the certification and validation parts of the project. The NEAN ground network was completed in May 1997 and airborne installations will be completed by the second half of 1997. • NEAP (sponsored by EC DGVII) The North European CNS/ATM Applications Project (NEAP) is a sister project to the NEAN, with the same participants. Using the infrastructure implemented in the NEAN, the NEAP will develop and demonstrate end-to-end (airborne and ground based) applications using the VDL mode 4/STDMA data link. The applications to be investigated in NEAP include: enhanced ATC surveillance (both while airborne and on the ground); uplinked support information for pilots, e.g. TIS data; uplinked differential GNSS corrections and integrity data. • FARAWAY (sponsored by EC DG XIII) The objective of FARAWAY is to investigate the enhanced operational performance of ground surveillance and aircraft navigation made possible through fusion of radar and ADS-B data. The Faraway project is co-ordinated by Alenia Spa, Italy and involves ATM service providers and airlines in Germany, Italy and Sweden. Initially three Alitalia MD-82 will be equipped with STDMA and cockpit display equipment and one ground station will be installed at Ciampino airport, Rome. The FARAWAY trials will run from October 1997 to March 1998. • MAGNET B (sponsored by EC DG XIII) The objectives of Magnet B are to develop GNSS1 user segments, to assess their capability to meet the most demanding aviation requirements and to evaluate the benefits that users can achieve from the integration of GNSS1 with a two-way data link. The Magnet B project is co-ordinated by Dassault Electronique, France and includes participants from Germany, UK, Norway, the Netherlands and Sweden. When practical trials start, it is expected that STDMA will be installed in an NLR aircraft in Holland and a base station also located there. • PETAL II (sponsored by EUROCONTROL) PETAL-II is a Eurocontrol project to investigate use of air-ground data link to perform real-time CPDLCs. Petal II is using the two-way data link capability of STDMA to provide this application. STDMA ground stations were installed at the Maastricht Centre and at the Eurocontrol Experimental Centre during March/April 1997. These trials programmes have provided early demonstration of the use of STDMA for ADS-B applications. There is the potential for integrating VDL mode 4 draft SARPs or ETSI standard compliant equipment with these trials in 1998 in order to provide an extensive test bed for the validation of the VDL mode 4 draft SARPs or ETSI standard. The early establishment of an ETSI standard is therefore desirable to provide a stable reference specification for development of this equipment. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.2.5 Summary | In deciding whether to standardize STDMA, the main evidence to be taken account of is therefore: • a requirement for STDMA has been expressed by some airlines; TR 101 130 V1.1.1 (1997-11) 19 • the ability of STDMA to offer a flexible communication system and, in particular, its ability to provide communication functions that are not currently supported by other VDL modes: • operation does not require ground infrastructure; • air-to-air communications, both with and without the presence of ground stations; • broadcast communications, both from ground and airborne users. • its ability to offer a solution for ADS-B applications; • progress of the STDMA standard within the ICAO forum is slow and there is no prospect of achieving an early standard. ETSI can provide a rapid route for standardization through its flexible standards development process: • making possible the establishment of a regional standard; • providing a benchmark for Commission funded trials activity; • promoting the development of the operational uses of STDMA. ETSI can therefore provide a path to promote a system being developed by European industry. In developing a possible standard, account should be taken of the development of other standards, notably: • Mode S squitter as an alternative ADS-B enabling technology; • VDL modes 2 and 3, taking opportunities, wherever possible, to unify the point-to-point functions of the three modes. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.3 Procedural issues | The production of an ETSI standard must take account of and, wherever possible, co-ordinate with other standards activities related to STDMA. The necessary activities identified in the study include: • co-ordination with standards being developed by ICAO AMCP, possibly through exchange of change requests; • taking account of emerging standards for ADS-B being produced by Radio Technical Commission for Aeronautics (RTCA); • co-ordination (and potential resource sharing) with VDL mode 4 Minimum Operational Performance Specifications (MOPS) being developed by European Organization for Civil Aviation Electronics (EUROCAE) WG-51; • information sharing with the Airline Electronic Engineering Committee (AEEC) in order to encourage AEEC to develop common interface standards; • co-ordination with groups developing STDMA-derived standards in other application areas, notably land and maritime. As a result of the analysis of the VDL mode 4 standards material upon which an ETSI standard would be based, it has been concluded by the STF that it will take approximately one year to produce an ETSI standard with the same boundaries as the current ICAO standard, followed by a further 6 months to produce an approval and protocol conformance specification. Since there is an urgent need to produce a standard to support trials activity carried out in 1998, a phased approach is recommend in which: • An initial TS (TS1) is produced which will define a system targeted at ADS-B applications and is the most natural extension of the STDMA system. It is estimated that such a TS could be produced after 6 months work. • A second TS (TS2) is produced which will define an enhancement to support point-to-point communication. This could be completed after a further 6 months work. TR 101 130 V1.1.1 (1997-11) 20 • A third TS (TS3) is produced to contain the approval and protocol conformance specification. An initial version of this would be produced after the first 6 months in order to support the trials use of the system defined by TS1. There would be a later extension produced after the completion of TS2 to define the full approval and protocol conformance specification for an operational system based on the combined TS1 and TS2. Note that it might be preferable to separate TS3 into two TS's, corresponding to conformance specifications for TS1 and TS2 respectively. • Once TS1 and TS2 are complete, it is proposed that they are submitted for formal approval to produce an EN. • It is proposed that the required work is carried out by two sub groups: • The first one will work on the physical layer; • The second one will work on the Logical Link Control (LLC) and network layer, which, in ICAO SARPs terminology includes the link layer (Media Access Control (MAC) sublayer, VDL mode 4 Specific Services (VSS) sublayer, Data Link Service (DLS) sublayer, Link Management Entity (LME) sublayer) and the subnetwork layer. It will be necessary to set up a liaison activity between the two groups to define the interfaces between the two layers and to liaise with other standards activities. This interface will probably evolve with the technical work within the two subgroups. More detail on the procedural issues relevant to the standardization of STDMA is contained in annex B. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.4 Standardization issues | |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.4.1 Introduction | The standardization of STDMA will have to follow the usual methodology: • define precisely the needs and the requirements that this standard will have to satisfy; • choose the technical solutions that will permit to offer solutions to the previously identified needs and requirements; • write a standard which clearly and unambiguously reflects the solutions adopted in the standard. The issues raised by the review process are discussed in this subclause. More detail on STDMA standardization issues is contained in annex C. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.4.2 The need for a clear system concept | |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.4.2.1 The need for clear interfaces | It is necessary to summarize the various functions that STDMA will have to support. Considering that STDMA is intended to support a lot of different interfaces, it may be necessary to organize the offered functions in relation to their main characteristics e.g. connection or connection-less oriented. The defined interfaces will have to be complete with not only the parameters of the requirement but also with the indications returned in case of failure or impossibility to satisfy the service. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.4.2.2 Clear boundaries for the system | Clear boundaries of the system will have to be defined. These boundaries concern the channel management and the way the functions are supported. One may distinguish between three main approaches to this problem: • To define a very simple communication system operating on a frequency. This frequency may vary in a given frequency bandwidth. • To define in addition to the previous communication system tools which may make it possible to manage various channels and to organize various functions sharing a same resource. TR 101 130 V1.1.1 (1997-11) 21 • To build a fully integrated telecommunication system capable of taking into account the various services requirement, their need of bandwidth and possible competition between these requirements. This problem has technical implications. The solution chosen will impact greatly on the architecture of the standard. Because the detailed requirements for services are not yet fully defined (see subclause 6.2.2), it is felt that the third of these approaches is unlikely to be practical. Instead it is recommended that a flexible standard is produced through a combination of the first and second approaches. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.4.2.3 Clear mode of operation in the given spectrum | The availability of the targeted spectrum has to be clearly identified as well as the possible constraints. One may be in various scenarios ranging from the exclusive use of the bandwidth to the coexistence or inter operation with other systems. As a priority, the standardization work should consider this issue and provide a resolution to it. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.4.3 The reliability issue | Concerning a telecommunication standard in the aeronautical field, STDMA will have to address the reliability issue. In the DGXIII Treaty 8 report [10], the functions described have an associated reliability requirements. These requirements will have consequences on the design of the STDMA standard. The reliability issue may be addressed throughout all the whole STDMA standard. For instance the reservation scheme which is a corner stone of the STDMA draft standard will have to address with this concern. Moreover the reliability issue imposes that a conformance testing specification will be worked out to ensure a proper operation of the system. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.4.4 Structure and presentation of the standard | The standard will be written in accordance with the chosen architecture and in accordance with ETSI best practice. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.5 Technical issues | This subclause describes the technical issues that will require investigation during the standardization process. There are four kinds of technical issues: • provision of procedures which are not yet specified in the draft standard; • the optimization of the parameters which are used in the standards; • the possible enhancement of the system; • the validation of performance with respect to requirements. The level of effort required to resolve these issues varies between topics. Some will require substantial effort. More detail on the technical issues relevant to the standardization of STDMA is contained in annex D. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.5.1 The missing procedures | |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.5.1.1 Secondary timing and positioning protocol | The timing and positioning recovery procedure when local UTC source fails is mentioned in the draft but is not specified. A clone of the GNSS procedure based on position broadcast of remote aircraft and tracking of their burst synchronization may suffice. The main difficulties are in the definition of hardware and software timing requirements for this purpose and in providing tests to demonstrate compliance. TR 101 130 V1.1.1 (1997-11) 22 |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.5.1.2 Channel management | The set of channel management procedures is very important. If the committee plans to include them in the draft standard then there will be the need to specify numerous protocols for dynamic channel assignment and for dealing with possible erroneous behaviours of the system in case of protocol failure. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.5.1.3 High priority reservation pre-emption protocol | This procedure is mentioned in the ICAO standard but not yet specified. Algorithms exist in case of a centralized protocol where a central agent rules the medium access. In a distributed protocol such as that used for STDMA, the problem might be less easy since the pre-emption poses problems in the case of contention between different priorities. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.5.2 The optimization parameters | Optimization of parameters in the ICAO standard should not require too much resource, provided the committee has the technical expertise to achieve it, or can rely on technical database and simulation tools. A number of STDMA simulation and modelling studies are currently in progress as part of the ICAO process and it is hoped that this expertise can be drawn upon to carry out the work. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.5.3 The system enhancements | |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.5.3.1 Superframe parameters | Some clarification is required of the superframe parameters. This is not expected to require significant resource to solve. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.5.3.2 Warming up procedures | The system recovery after a long channel failure or when an aircraft first enters STDMA coverage may be considered too slow, although this may not prove to be operationally significant. A study must be carried out to investigate this issue and possibly to propose an enhancement to the standard. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.5.3.3 Slot reuse and Co-Channel Interference (CCI) conditions | The CCI condition for slot reuse may be incomplete since it considers only one transmitter per slot, while several transmitters are possible. In this case the CCI conditions may need to be adapted and the resulting broadcast condition may be more complicated. CCI conditions based on Received Signal Strength Indication (RSSI) power measurement instead of position estimate may prove to be an interesting alternative to the currently proposed scheme. It may be possible that CCI conditions and broadcast conditions could be simplified and merged to a single condition based on RSSI with negligible performance degradation. The issue of slot re-use therefore needs substantial study and validation through modelling. Once again, existing simulation work could be used to assist this work. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.5.3.4 Distributed clock software synchronization | A timing mechanism that does not rely on an external source might be investigated to increase the integrity of the system. Software synchronization mechanisms provide interesting results but require a priori a communication whose reliability is independent of clock synchronization, see references [11], [12], [13]. Investigations to find possible less demanding protocols are recommended. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.5.3.5 Reaction to long jamming periods | Procedures to deal with jamming and other persistent noise events need to be investigated. TR 101 130 V1.1.1 (1997-11) 23 |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.5.4 Viability check and performance study | |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.5.4.1 Modulation choice | STDMA is sufficiently flexible to support two modulation types. The choice of modulation type may need further study. However for ADS-B and Differential GNSS (NABS) applications, it is expected that GFSK will prove to offer the best system performance (see reference [5]). A study needs to be carried out to investigate the relative advantages and disadvantages of the possible modulation types and to make recommendations for the best scheme as a function of service provided. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.5.4.2 The slot reuse and broadcast reliability | The periodic broadcast is the most demanding function in STDMA. The access protocol needs to guarantee, with sufficient reliability, the performance of the network in the case of heavy traffic, or any other critical conditions against which STDMA needs to be tailored. For example the E-TDMA study (see reference [10]) outlines a worst case en-route situation of 570 aircraft in a radius of 160 nmi, each aircraft broadcasting its position every six seconds. In this very case it is necessary to check that the broadcast transmissions are received with enough range and enough reliability. A modelling and simulation effort is required and an example is given in annex D. Reuse factor, average reception area and reliability estimates are the parameters which should be derived as a result of this study and the result should be used to further develop the system concept including consideration of number of channels required, use of GSCs etc. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 6.5.4.3 Remote and hidden node management | One should check under which conditions an aircraft could be visible for a second station and not for a third one. One case is the ground effect on ground stations. The committee should investigate the way to cope with this problem which may affect ground control. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 7 Proposed work plan for STDMA standardization | |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 7.1 Introduction | This subclause presents a proposed work plan for STDMA standardization. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 7.2 Key issues and recommended approach | The European Commission has indicated that it is desirable to produce an EN for STDMA. This is desirable because it adds force of European Law to the standard and will have most influence on the ICAO process. However, the procedure for achieving an EN will result in delay which may not enable the full impact on the ICAO process to be realized. As described in subclause 6.3, the STF recommend that the development work is divided into three TSs. The first two of these TSs would then be submitted to the full voting procedure to produce an EN. This approach results in the early availability of TC approved standards which can then be converted to EN status in the fullness of time. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 7.3 Organization of the work | The recommendation derived from this study is to carry out the development work using the following procedure: • build a set of functions and associated requirements that STDMA will offer; • develop a clear system concept and architecture for the system; • break the committee in charge of the standardization into two subgroups: The Transmission Techniques Group and Protocol Design Group. The Transmission Techniques Group should study the following points: • determining an appropriate choice of modulation; TR 101 130 V1.1.1 (1997-11) 24 • specification of parameters concerning power limits, switching times, etc.; • ensuring the interoperability with VDL mode 2 and other media such as ACARS VHF, voice VHF, navigation aids etc.; • defining antenna requirements to ensure a nearly isotropic radiation; • defining requirements and suitable techniques for timing synchronization. The Protocol Design Group should study the following points: • functions to be offered by STDMA and related interfaces; • system concept definition, channel management and function management; • timing synchronization for example, what form of distributed algorithm will be used to synchronize clocks); • slot reuse algorithms; • reliability issues in the STDMA functionality. The two sub-groups will focus on the production of the following deliverables: • An initial TS (TS1), which will define a system targeted at ADS-B applications and is the most natural extension of the STDMA system. It is estimated that such a TS could be produced after 6 months. • A second TS (TS2), which will define an enhancement to support point-to-point communication. This could be completed after a further 6 months. • A third TS (TS3) to contain the approval and protocol conformance specification. An initial version of this would be produced after the first 6 months in order to support the trials use of the system defined by TS1. There would be a later extension produced after the completion of TS2 to define the full approval and protocol conformance specification for an operational system based on the combined TS1 and TS2. Note that it might be preferable to separate TS3 into two TSs, corresponding to conformance specifications for TS1 and TS2 respectively. In addition, the committee should consider whether a TR deliverable detailing the functions and facilities of the system and the services and applications to be supported should be produced as an aid to explaining the purpose, interfaces and boundaries of the system. The TR should also specify the type approval requirements for the system so as to define the level of conformance testing necessary. Such a document could be produced at the start of the recommended work as a means of providing a reference work for the study. The committee will need to provide control and co-ordination of the overall activity. This will include: • defining the terms of reference of each sub-group and ensuring that there is no duplication of effort; • providing editorial control over the production of the TS document; • providing liaison support with other standards bodies, notably ICAO, EUROCAE and International Maritime Organization (IMO). Note that technical constraints may impose that TS1 and TS2 be within the same document (see subclauses B.3.5 and C.2.2). As was discussed in subclause 6.2, account should be taken of the development of other standards, particularly VDL modes 2 and 3, taking opportunities, wherever possible, to unify the point-to-point functions of the three modes. It is recommended that a review of the current state of standardization of these standards is carried out prior to starting work on TS2 in order to decide on the need for TS2 and the required content of this part of the standard. TR 101 130 V1.1.1 (1997-11) 25 |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 7.4 Timescales for the work | It is estimated that it will take 6 months to one year to complete the functional specification work. An additional 8 months will be necessary to perform the radio approval and protocol testing. An initial timing chart is provided in figure 3 showing approximate timescales. It is recommended that a detailed planning activity is carried out at the start of the study. It may be necessary to increase the time taken to carry out the production of TS1, leading to slippage of the delivery date. This will depend both on available resource and on the time taken to complete the first task (production of the TR). If additional time is necessary, it may be possible to reduce the time taken to produce TS2 so as to keep to the overall planned timescales. The detailed planning must also take account of the timing of TC ERM meetings, particularly as a decision to proceed with the production of TS2 will have to be taken by that committee after the completion of TS1. ID Task Name 1 Clarify functions and facilities to be supported by STDMA 2 Technical Report 3 Develop and specify broadcast functions for STDMA 4 Technical Specfication 1 (TS1) 5 Develop and specify point to point functions 6 Technical specification 2 (TS2) 7 Develop conformance tests to support trials activities 8 Initial version of TS3 9 Develop conformance test to support operational system 10 Technical specificaiton TS3 11 Voting process for EN based on TS1/TS2 12 EN for STDMA 28-01 17-06 02-12 02-12 14-07 14-07 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 1st Quarter 2nd Quarter 3rd Quarter 4th Quarter 1st Quarter 2nd Quarter 3rd Quarter 4th Quarter Figure 3: Initial study timescales (assuming start date 01/01/98) |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 7.5 Estimated resource requirements | According to the previous organization of the work and taking into account annex B, C and D for further refinement, one can estimate that a committee must be set up to cover a period of 20 months (in accordance with the timescales set out in figure 3). According to the initial timescales, this work can be divided between around 12 months for the functional specifications and around 8 months for the conformance testing issue. The exact number of meetings necessary is difficult to assess, but it could be assumed that the committees would have to meet at least every 2 months. Since it is probably unlikely that sufficient resources can be found within ETSI to enable the work to be carried out wholly by a committee, it is recommended that an STF is used to: • carry out the bulk of the required work and provide specialist expertise in air traffic management and STDMA (estimated effort required is 14 man months, dependent on the level of expertise available within the committee); • support the conception of the whole system including modelling and simulation of various possible approaches, dimensioning, ensuring reliability and performances of the system (estimated effort required is 6 man months); • to carry out most of the work on the radio approval and conformance testing specification (estimated effort required is 10 man months). Overall, therefore, it is recommended that over an elapsed period of 20 months, the effort required to carry out the work is estimated to be 30 months of contracted STF effort. |
c8f0bbee7e14c2dab1d253396507cc82 | 101 130 | 8 Conclusions and recommendations | In accordance with CE Mandate M/239, a Special Task Force has carried out a study on the feasibility of standardizing self organized time division multiple access (STDMA) system requirements. The available specification material for STDMA has been analysed to see what actions ETSI should take to: • transfer the technical specification into a standard; TR 101 130 V1.1.1 (1997-11) 26 • complete all the required interfaces; • identify areas needing particular attention; • define the appropriate time schedule for any future work. The following conclusions have been reached as a result of carrying out this review: • STDMA appears to provide a flexible data link that has the potential to act as an enabling technology for a wide range of future Air Traffic Management applications. In particular, the system has been optimized to support ADS-B applications. • There is a market requirement for the system. • Although there are rival candidate systems for ADS-B, notably a system based on mode S squitter transmissions, there is, as yet, no international consensus on the eventual choice of data link for ADS-B and it is not certain when such a consensus could be reached. • STDMA has a number of unique functions that differentiate it from other VDL standards, notably: • operation with no ground infrastructure; • air-to-air communications, both with and without the presence of ground stations; • broadcast communications, both from ground and airborne users. • Current trials activity urgently require the establishment of a standard in order to promote further understanding of the operational uses of STDMA. • The current progress towards standardization of STDMA (known as VDL mode 4) within the ICAO forum is slow and there is no prospect of achieving an early standard. • The current ICAO standards provide a reasonable starting point for ETSI standards but the review has identified a number of presentational and technical issues that must be resolved before a standard can be produced. These issues include the need to: • define clear interfaces, system boundaries and the mode of operation for the system; • change the presentation of the standard to conform with ETSI best practice; • define procedures that are currently absent from the standard including timing, channel management and pre- emption protocols; • optimize certain parameters; • consider possible system enhancements including alternative slot re-use algorithms based on RSSI and resistance to jamming and other interference sources. • carry out system viability checks including modelling of broadcast delivery reliability and remote and hidden node management. The STF makes the following recommendations: • A standardization activity for STDMA should be started within ETSI based on the existing ICAO standard. The aims of the standardization activity should be to: • respond to market need; • support the development of the STDMA system, particularly the functions provided by system, namely broadcast functions, air to air functions and operation without ground infrastructure, that are currently not supported by any other VDL modes and which enable STDMA to support a wide range of applications and to provide a data link in core continental, oceanic and low density continental airspace; • provide a stable standard for trials assessment of ADS-B applications; TR 101 130 V1.1.1 (1997-11) 27 • provide extensions to the standard that are capable of supporting point-to-point communication. • The deliverables for such an activity should consist of: • An initial TS (TS1), which will define a system targeted at ADS-B applications and is the most natural extension of the STDMA system. It is estimated that such a TS could be produced after 6 months. • A second TS (TS2), which will define an enhancement to support point-to-point communication. This could be completed after a further 6 months. • A third TS (TS3), which will contain the approval and protocol conformance specification. An initial version of this would be produced after the first 6 months in order to support the trials use of the system defined by TS1. There would be a later extension produced after the completion of TS2 to define the full approval and protocol conformance specification for an operational system based on the combined TS1 and TS2. Note that it might be preferable to separate TS3 into two TS's, corresponding to conformance specifications for TS1 and TS2 respectively. • An EN based on TS1 and TS2. The voting process for this EN will start once both TS1 and TS2 are complete. • In addition, consideration should be given to the production of a TR deliverable detailing the functions and facilities of the system and the services and applications to be supported. This would be useful as an aid to explaining the purpose, interfaces and boundaries of the system. The TR should also specify the type approval requirements for the system so as to define the level of conformance testing necessary. Such a document could be produced at the start of the recommended work as a means of providing a reference work for the study. • The committee in charge of the standardization should be divided into two subgroups : The Transmission Techniques Group and Protocol Design Group. • Overall, therefore, it is recommended that over an elapsed period of 20 months, the effort required to carry out the work is estimated to be 30 months of contracted STF effort. • Development of STDMA should take account of activities in other standards bodies. In particular, account should be taken of: • VDL mode 2 standards and the emerging VDL mode 3 standards within ICAO. Some functions provided by STDMA are also provided by VDL mode 2 and mode 3 and any further development of these functions within STDMA should be fully justified in terms of operational advantage and user benefit. It is recommended that a review of the current state of standardization of these standards is carried out prior to starting work on TS2 in order to decide on the need for TS2 and the required content of this part of the standard; • emerging standards for ADS-B being produced by RTCA and also alternative ADS-B system solutions such as mode S squitter; • the need for co-ordination (and potential resource sharing) with VDL mode 4 MOPS being developed by EUROCAE WG-51; • the potential benefits of information sharing with AEEC in order to encourage AEEC to develop common interface standards; • the need for co-ordination with groups developing STDMA-derived standards in other application areas, notably land and maritime. TR 101 130 V1.1.1 (1997-11) 28 Annex A: Design principles of STDMA A.1 Functions and traffic requirement addressed by STDMA Typical operational scenarios to be supported by STDMA mix broadcast applications (and, in particular, ADS-B applications) with point-to-point applications (CPDLC etc.). ADS-B applications are expected to be the main traffic contributor to STDMA. ADS-B requires that aircraft position parameters are periodically broadcast to other aircraft (air-air) and to ground stations (air-ground). The ADS-B operational environment places special requirements on the design of a multiple access technique to be employed. • Firstly, aircraft ADS-B transmitters are highly mobile such that the environment around any given surveillance receiver is rapidly changing. • Secondly, it is necessary that for air-air ADS-B applications, the system should be able to operate autonomously without a central controller in many environments, including oceanic and low density continental airspace. • Finally, ADS-B must appear seamless to an ADS-B surveillance receiver; i.e. an airborne sensor must always be able to hear all nearby aircraft reports regardless of their location with respect to any "cell" or "sector" boundaries. Point-to-point functions are also important requirements in order to support ATN protocols and these could be supported by STDMA if the point-to-point functions defined within the ICAO VDL mode 4 draft SARPs are incorporated into the ETSI standard. However, its performance will need to be compared with that offered by other VDL standards such as modes 2 and 3. Note that in order to achieve optimum use of the data link, it may be necessary to partition the different applications on to different channels and to use different parameters (e.g. guard time, modulation scheme, slot length etc.) for each channel. Obviously, this increases the system complexity and is probably not desirable if it can be avoided. Because ADS-B requirements dominate the expected operational scenario, it is expected that STDMA will be optimized with respect to this service. Also, because STDMA could support point-to-point communication (particularly if such support is on a separate channel), it is expected that it will provide a potential solution for end-to-end applications. A.2 STDMA system concepts and frequency allocation A.2.1 Introduction The system concept should take account of the frequency shortage in Europe and elsewhere. A concept based on distinct coverage cells relies on different frequency cells for data link activity and will probably have to use spectrum currently used for voice communication. There will therefore be difficult transition issues to solve. A possible solution in which a frequency for data link is assigned to each air traffic sector is probably not practical in the early days since there will be insufficient spectrum (it is possible that the migration of voice to 8,33 kHz channels may free up some spectrum in the longer term and if not limited to upper airspace only). A better approach may be to define a large region served by the same frequency and which operates near to capacity in the early days, and to reduce the size of that region as the data link load increases and as more spectrum becomes available. Alternatively, use of a system that allows re-use of the same channel may be a better, more efficient approach. This subclause explores these issues in more detail. TR 101 130 V1.1.1 (1997-11) 29 A.2.2 System concept CFP is a VHF coverage concept which is organized into distinct coverage cells, such that there is no overlapping coverage between cells using the same frequency. RFP is a VHF coverage concept which allows the same frequency to be used on adjacent cells, taking advantage of capture effect on packet reception. The proposed STDMA concept uses RFP and does not require specific CFP. Note, however, that STDMA does not preclude the use of CFP in its channel management architecture. CFP has the advantage over RFP that it allows a very simple and centralized channel access management around every base station. RFP needs more sophisticated distributed channel access which takes into account interference from remote stations and hidden nodes. However the RFP concept presents numerous advantages over CFP as listed below: • CFP consumes greater frequency resource since strict cell mapping would require at least four distinct frequency sets (experience with GSM suggests that at lest six distinct frequency sets would be required). In theory RFP would need only one set of frequencies. Furthermore frequency planning on empty areas (i.e. without ground stations) would be difficult to manage. • In general CFP is difficult to manage in a dynamic configuration. For example as air traffic load increases it will be difficult to add new frequencies in each cell or to reduce the size of cells. Adding extra frequencies in RFP is trivial. Note that, in RFP, a higher load on the same frequency may simply lead to smaller link range which may be tolerable if the available range does not drop below critical values. • RFP does not need a channel hand-over protocol between cells. In CFP, handover is a potentially weak part of the system since channel misallocation could disrupt the communication link. STDMA can use the same channel on two adjacent cells, thus neighbouring ground stations will be able to monitor the air traffic beyond their own cell boundary. The system concept can have an important impact on the data link protocols. For example: • Aircraft moving from one ground station to another in an RFP concept, require co-ordination of coverage in overlapping regions to ensure that there is no channel contention. Such co-ordination could be provided by a ground network (with resultant networking design, implementation and running costs) or (preferably) could be provided by link management protocols defined as part of the data link standard. • Aircraft moving from one cell to another in a concept using distinct coverage cells also require co-ordination. For point-to-point communication, this involves maintaining peer-to-peer links and is achieved by ATN protocols. The situation is more difficult if ADS-B applications are being supported, since all aircraft in the overlapping region needs to receive transmissions from all other aircraft in the region, some of which will be operating on different channels. In this case, the data link protocols need to support parallel operation on more than one channel. Therefore ADS-B applications suggest the use of global channels, since otherwise, a non-trivial co-ordination mechanism will be needed to maintain a full surveillance picture for all users in the transition region from one frequency to another. In this case, RFP management, which is supported by STDMA, is expected to be a key for such an application. ADS-B ground-air applications could be implemented using a CFP concept with the transition from one frequency to another controlled by the ground system. However, since ADS-B could be used to provide the ground based surveillance service, the control mechanism would have to be of the highest integrity and could result in costly ground networking solutions. Furthermore, local constraints may impose CFP for VDL. In this case, the STDMA concept does not preclude the use of any particular channel management technique, with or without CFP in its channel architecture. STDMA will provide the elements needed for the implementation of the preferred management scheme. A.2.3 Frequency allocation The bands currently allocated to the aeronautical service by ITU and potentially available for data link are: • VHF with 50 and 25 kHz channelization (currently used for navigation and VHF communications). • L-band with 1 and 8 MHz channelization (currently used for navigation and secondary surveillance radar). TR 101 130 V1.1.1 (1997-11) 30 Both of these bands have advantages and disadvantages for use by an ADS-B system: • Power requirements: Propagation path loss is proportional to the transmit frequency. This generally allows systems operating in a lower frequency band to use less transmit power than a comparable system at a higher frequency. Since the transmitters are airborne-based for this service the advantage of lower power could be an important consideration, particularly for the General Aviation (GA) market. • Spectrum availability: The VHF band is constrained due to current use world-wide. Given the current constraints, a system operating in this band must operate within the channelizations of 50 or 25 kHz to "fit in" amongst current assignments. Currently VHF communications channels are in great demand. Therefore, initially finding and co-ordinating the international use of one or more of these channels solely for ADS-B may be difficult. However, depending on the actual installed implementation of navigational aids, there may exist the possibility to intermix ADS-B channels in that portion of the VHF band. In addition to the old 112,0-136,0 kHz band assigned by ITU for aeronautical functions there are also new 25 kHz channels assigned to aviation by ITU in the 136,0-137,0 kHz band of which four of them have been designated by ICAO for data link functions. Within L-band it may also be possible to co-ordinate a global channel due to the nature of the current ICAO band plan for Distance Measuring Equipment (DME) channel pairing, where one channel of the DME pair overlaps the SSR portion of the band. Since that DME channel pair cannot be used because of interference with secondary surveillance radars, then one channel of the pair is for the most part unused and could provide a clear channel. STDMA operates in the VHF band and therefore offers the potential for a low power and hence low cost data link. However, problems of VHF channel congestion will have to be solved to allow a suitable allocation to be made for its use. A.3 STDMA modulation scheme and physical layer STDMA is based on high speed (10 to 30 kbps) channels in order to satisfy the load needed by the traffic requirement. Among its optional scheme there is a modulation scheme GFSK at 19,2 kbps, which is resistant to noise (Signal to Noise Ratio (SNR) of 7 dB) which is needed for high link reliability. Furthermore it allows a good frequency reuse and is particularly well adapted for RFP. In particular ADS-B applications would need the use of a global channel with the following requirements: • use of a modulation scheme that can tolerate CCI from other users on the same channel (this points towards lower data rate); • a relatively large guard time to enable a user to receive transmissions from distant users (resulting in a reduced channel capacity). Point to point communications in ATN protocols would require: • use of a modulation scheme that maximizes data rate (this may points towards schemes that have relatively poor CCI performance, which limits frequency reuse but ATN protocols could be implemented in CFP); • a guard range targeted at the expected coverage range for each cell (probably results in a smaller guard time than used for a concept that uses a global channel and therefore is a more optimum use of the channel leading to higher capacity). STDMA proposes an alternative modulation scheme which is D8PSK at 31,5 kbps. A.4 STDMA Access protocols General access protocols for VDL need to be derived: • VDL mode 2 uses random access on a shared channel for ground-air communication. • VDL mode 3 uses ground controlled TDMA for ground-air communication. TR 101 130 V1.1.1 (1997-11) 31 • STDMA offers a slot reservation process which is particularly suited to meet the requirements for periodic broadcast of short packet as specified in ADS-B. The channel access of STDMA is distributed which is important to support RFP and does not need a centralized management (while not precluding it). It allows the system to work without ground control (opening up the possibility of air-air point-to-point communication). STDMA therefore offers a flexible channel access scheme that offers enhanced efficiency through the use of reservation protocols and distributed access. A.5 STDMA interoperability with other standards STDMA, as defined in VDL mode 4 draft SARPs, will mandatorily be backward compatible with ICAO VDL mode 2. TR 101 130 V1.1.1 (1997-11) 32 Annex B: Procedural issues B.1 Introduction This clause describes a number of procedural issues relevant to standardization of STDMA: • the relationship to other standards; • the selection of a standardization route. B.2 Relationship to and co-ordination with other standards This clause describes the various standardization activities relevant to STDMA. B.2.1 ICAO - International Civil Aviation Organization Global standards for aviation are published as annexes to the Chicago Convention. In annex 10 SARPs for Aeronautical Telecommunications are published, where VDL is one part. The SARPs material is prepared by ICAO panels of experts. Several different panels can be identified which have an interest in the area of STDMA: 1) AMCP - Aeronautical Mobile Communications Panel: This panel has already published standards for satellite communication and is currently working on High Frequency (HF) and VHF Digital Link (VDL) standards. Draft VDL mode 4 SARPs have been developed by the AMCP. The documentation consists of: • draft SARPs version 5.4 [1] and [3]; • draft Manual version 1.0 [2]. There was reasonable expectation of published standards for VDL mode 4 in 2000 on the assumption that a suitable panel meeting is held in 1999). However, ICAO has just announced that the next panel meeting (AMCP/5) will be held in Spring 1998. This is likely to result in further delay for VDL mode 4 since acceptance would now have to wait until the next panel meeting (AMCP/6) which is now more likely to be held later than 1999 (probably 2000). 2) ADSP - Automatic Dependant Surveillance Panel: This sets requirements for ADS applications and the deliberations of this panel will therefore impact on the potential surveillance uses of STDMA. 3) GNSSP - Global Navigation Satellites Systems Panel: This sets requirements for GNSS technology and applications and the deliberations of this panel will therefore impact on the potential navigation uses of STDMA, particularly with respect to the use of STDMA for differential correction uplinks. GNSSP is expected to choose a specific data link in the near future for GNSS augmentation signals - STDMA is a candidate for such a data link. 4) AWOP - All Weather Operation Panel: This sets requirements for applications to support poor weather conditions (i.e. poor visibility on approach and taxiing) and the deliberations of this panel will therefore impact on the approach and ground movement uses of STDMA. 5) SICASP - SSR Improvements and Collision Avoidance Systems Panel: This sets requirements for surveillance systems and applications and the deliberations of this panel will therefore impact on the potential surveillance uses of STDMA, in particular the use of ADS-B for Airborne Separation Assurance (ASAS). The next panel meeting, SICASP-7, will be held in 2000. Sub-group 2 looks at ACAS and ASAS. TR 101 130 V1.1.1 (1997-11) 33 It is important that standards activities undertaken by ETSI are co-ordinated wherever possible with those carried out by ICAO to ensure that standards for mode 4 do not diverge. However, because the ETSI process is more efficient, ETSI can provide support to the ICAO process by providing an early standard which can later become adopted as ICAO SARPs. B.2.2 RTCA - Radio Technical Commission for Aeronautics RTCA is an organization formed by the US industry. Special Committees are formed to carry out particular tasks, e.g.: • SC172 - VHF Digital Link (VDL); • SC186 - Automatic Dependence Surveillance Broadcast (ADS-B). RTCA is publishing Minimum Aviation System Performance Standards (MASPS) and MOPS for ADS-B. ETSI needs to take account of these emerging standards for ADS-B in optimizing the standards for STDMA. B.2.3 EUROCAE - European Organization for Civil Aviation Electronics EUROCAE is the European equivalent to the RTCA in the US. To some extent these two bodies co-operate and share their results. Working Groups are formed with special tasks: • WG41 is working on advanced SMGCS. • WG47 is working on MOPS for mode 1 and mode 2 VHF data links. • WG51 is working on ADS-B. EUROCAE MOPS based on VDL mode 4 draft SARPs which will only deal with those parts of the standard necessary to support ADS-B applications, although it has not yet been determined how to (or even whether it will be possible to) make a suitable partition of ICAO SARPs to reflect this. The issue of partition also needs to be addressed for the ETSI work (see subclause C.2.2). Note that MOPS depend on SARPs but could be available earlier (therefore MOPS will really be draft only until SARPs is published). • WG 53 is working on the definition at a system level of the future data link applications. Decisions made in this group could influence the targeted availability of STDMA. A key issue is to align the development of ETSI standards for mode 4 with the MOPS development process, so a to prevent duplication of work. It is expected that draft MOPS will be available in mid-1998 and hence this would fit reasonably well with the development of the first ETSI TS proposed in subclause B.3.5. B.2.4 AEEC Airline Electronic Engineering Committee The AEEC is an airline organization which is defining standards for how equipment physically should be designed. Purpose is to ensure common interface standards at the lowest level (form, fit and function). AEEC is organized and driven by ARINC, who are a leading airline operators communications service provider. It is important to inform AEEC of ETSI standards activities in STDMA and to encourage them to provide suitable AEEC standards. TR 101 130 V1.1.1 (1997-11) 34 B.3 Selection of standardization route B.3.1 ETSI documents and procedures ETSI can produce three different kinds of standardization documents: • Technical Specification (TS); • ETSI Standard (ES); • European Standard (Telecommunications series) (EN). These documents correspond to three different procedures for the elaboration of the document: • The technical specification reflects the work of a technical committee and its agreement on the obtained work. The document is formally approved after work in sub technical committee by the technical committee. • The ETSI specification is obtained from a technical specification if this ES is voted by ETSI members. ETSI has a formal voting procedure which takes 16 weeks. This vote is only between ETSI members. • A TS or an ES can become an EN. The document has to follow the public enquiry procedure. After this period of examination the document receives comments explained by countries represented at ETSI. These comments are then treated during a resolution meeting which provides a formal document which is then proposed to the vote. After the result of the vote the document is either accepted or rejected. It is understood that the EC wants an EN and this is desirable because it adds force of European Law to standard and will have most influence on ICAO process. However, the procedure for achieving an EN will result in delay which may not enable full impact on ICAO. The members of STF 109 believe that the type of document (TS, ES, EN) which is to be started by ETSI around STDMA is a committee choice. However, a set of recommended deliverables are presented in subclause B.3.5. B.3.2 Proposed approach to the work In the following we try to divide the remaining work to do to obtain an STDMA specification and to give recommendations about the amount of effort that is required to obtain this specification. As usual in a radio telecommunication system we can divide the standardization activity in two sub groups which can work quite independently: • The first one will work on the physical layer; • The second one will work on the LLC and network layer, which, in ICAO SARPs terminology includes the link layer (MAC sublayer, VSS sublayer, DLS sublayer, LME sublayer) and the subnetwork layer. It will be necessary to set up a liaison activity between the two groups to define the interfaces between the two layers and to liaise with other standards activities. This interface will probably evolve with the technical work within the two subgroups. The STDMA specification will require, as described in the standardization issues in subclause B.3.5 an attached specification concerning the radio type approval and conformance testing. STF 109 advise that it is possible to carry out this work after the STDMA functional specifications are complete and the results could be gathered into a separate document (probably a TR or EG). However the work for the technical specification needs to take into account the need to be able to test the radio protocols. This criteria will be very important in defining the technical specification. B.3.3 Physical layer work The work for the physical layer seems to be the simpler. Mainly we find the following remaining points to study: • determining an appropriate choice of modulation; TR 101 130 V1.1.1 (1997-11) 35 • specification of parameters concerning power limits, switching times, etc.; • ensuring the interoperability with VDL mode 2; • defining antenna requirements to ensure a nearly isotropic radiation; • defining requirements and suitable techniques for timing synchronization. The following effort is foreseen for the tasks involved in this work: • For the modulation choice, the missing figures and the check of the interoperability with VDL mode 2, one can foresee that these issues can be solved with two or three sub-group meetings if convenient inputs are given before the meeting. • The definition of antennas to ensure a nearly isotropic radiation can probably be solved in one sub-group meeting. • The timing synchronization is the most difficult part of the physical layer. It can be estimated that this work will require between two to four meetings depending on the complexity of the choices . Moreover it is possible that if a synchronization algorithm based on packets exchange is used then this will mix deeply the physical layer and the network layer. These tasks could be carried out independently of each other. B.3.4 LLC and network layer approach The work for the network layer is more important and will depend on the technical choices that will be done during the committee work. The following points need to be covered: • functions to be offered by STDMA and related interfaces; • system concept definition, channel management and service management; • timing synchronization for example, what form of distributed algorithm will be used to synchronize clocks); • slot reuse algorithms; • reliability issues in the STDMA functionality. The following effort is foreseen for the tasks involved in this work: • We can foresee that the discussion about the functions to be offered by the STDMA standard may be solved in two meeting with proposals before the first meeting. The proposal will be discussed during the first meeting. Another meeting may be necessary if there is a lack of expertise in the aeronautical fields. With the selected functions to be offered by STDMA we have to attach precise requirements. These requirements will deal with bandwidth, access delays, reliability . It could be useful if at the end of this work the committee will be able to define a model for the traffic that STDMA will have to support. This model will be used to set up objective comparisons. • The clear definition of the functions to be offered with their attached requirements will be the basis of the definition of the system concept. The channel management and quality of service management is deeply linked with the definition of the system concept. This work may call for modelling, simulations, objective comparisons between various approaches. There may need to be a formal proof to ensure proper operation of the system. One can evaluate that this work will require between two to six meetings. It is likely that the time required to carry out this work, which will probably require simulation and modelling, could be reduced with the help of an STF. • The timing synchronization is an independent work that can be done independently of the previous two main tasks. Depending on the solution that will be chosen we can evaluate the amount of work between one to four meetings. • The slot reuse is also an independent task. This work will require simulation and analytical models. This a very technical task where the help of a STF could be useful. The duration of this work can be evaluate between two to four meetings. TR 101 130 V1.1.1 (1997-11) 36 • Reliability issues can be considered as a general task that is to be considered in all the procedures that will be defined. We can notice that in the physical layer the schedule is less tied than in the network layer. We have interaction between the work in the physical group and the work in the network group especially one the following points: timing synchronization, slot reuse and modulation choice. According to this division of the work one can see that the critical path is within the protocol work and mainly on the system concept definition and the corresponding procedures to offer the functions. B.3.5 Phased approach to standard production and proposed timescales for the work Actually it is possible that a further division in the standardization work can be introduced. This split comes from a division between the functions offered by STDMA. These parts will be: • a system targeted at ADS-B applications; • enhancement of the system to support point-to-point communication and compatibility with modes 2 and 3. The verification that such a split is possible should be an early sub-group task. Deliverables associated with each part are discussed below. An initial TS (TS1) will define a system targeted at ADS-B applications and is the most natural extension of the STDMA system. The TS would provide a stable baseline for equipment to be produced in mid-1998 to support trials activity. The offered functions will be as follows: • Autonomous scenario, consisting of: • autonomous aircraft scenario (i.e. no ground station); • periodic broadcast protocol; • incremental broadcast protocol; • combined incremental/periodic broadcast protocol; • random access protocol; • autonomous net entry; • Ground station functions, consisting of: • directed request protocol; • procedures for ground directed control. A next TS (TS2) will define an enhancement to support point-to-point communication. Although derivation of the necessary functions could be based around the current draft ICAO SARPs [1, 2], the timescales for production of this TS (say late 1998) would allow a detailed study of the usage of a separate data communication channel and the compatibility with modes 2 and 3 to be examined in detail. Possible modifications to optimize the mode 4 standard for data communications could be proposed at this stage. The proposed functions of TS2 would be: • End-to-end VSS protocols, consisting of: • remaining VSS protocols to provide end-to-end VSS functions; • unicasted protocol; • information transfer protocol; TR 101 130 V1.1.1 (1997-11) 37 • DLS protocols, consisting of: • build the DLS protocols on top of the VSS protocols in order to provide ISO 8208 [19] compatible data link functions; • DLS protocols will be restricted to ground/air communication only; • connect to ATN using the existing mode 2 model functions wherever possible; • Subnetwork layer functionality, consisting of: • ATN connection; • Air-air end-to-end communication; • Multi-cell support. It is proposed that TS1 and TS2 are then submitted as an EN, for agreement some time in 1999. It is also possible that these two documents can not be separated. In that case a single document containing the whole system could produced for late 1998. This document could be submitted as EN, for agreement some time in 1999. The standard concerning the functional specifications will have to be followed by another specification for radio type approval and protocol conformance specification. These conformance specifications can be gathered in an ETSI document TS3. The radio type approval and the protocol conformance testing can be worked out independently. This work is extremely technical and does not generally generate long debate. One can estimate this testing work between six to eighteen months depending on the complexity of the network protocols and the offered support to do this work. Companies can support this work but the past shows that this work is completed more rapidly by a dedicated STF. It can be foreseen that this conformance document will be available in late 1999 and could submitted as EN for agreement in 2000. It can be noticed that the conformance testing of corresponding to TS1 can be started at the end of TS1 allowing to have earlier TS1 with a conformance specification. If the alternative of two TSs (TS1 and TS2) is adopted, one may envision to have two TS for the conformance testing related respectively to TS1 and TS2. In addition, the committee should consider whether a TR deliverable detailing the functions and facilities of the system and the services and applications to be supported should be produced as an aid to explaining the purpose, interfaces and boundaries of the system. The TR should also specify the type approval requirements for the system so as to define the level of conformance testing necessary. Such a document could be produced at the start of the recommended work as a means of providing a reference work for the study. Account should be taken of the development of other standards, particularly VDL modes 2 and 3, taking opportunities, wherever possible, to unify the point-to-point functions of the three modes. It is recommended that a review of the current state of standardization of these systems is carried out prior to starting work on TS2 in order to decide on the need for TS2 and the required content of this part of the standard. TR 101 130 V1.1.1 (1997-11) 38 Annex C: Standardization issues C.1 Introduction The standardization of STDMA should follow the rules that are common to the standardization of telecommunication systems: • First, one has to define the need and requirements that this standard will satisfy (for example, to avoid the lack of capacity and services on the present systems). This exercise is important and not so easy because the communication system based on the standard will probably have a long lifetime. Therefore the need and requirement shall take into account not only currently existing applications but also to forecast possible air traffic increases and the coming of new applications. • Second, one has to determine and define how these needs and requirements will be satisfied. Two areas must be investigated: • The first one concerns the physical layer: how data will be coded, modulated and received. • The second one concerns the network architecture: how the mobile stations will organize their access to the medium and how the services will be supported. Sharing a communication medium organized in multiple access is a problem which has received intensive attention for more than 20 years and there are numerous classes of solutions which are well studied. The scheme proposed for the standard is well documented in references [1] and [2] and belongs to the family of slotted reservation access. Since the radio spectrum resource is scarce, it is important to identify immediately the allocated frequencies and bandwidth and under which conditions they will be used. Last but not least, the protocols adopted in the standard need to be written in a way that removes all kinds of ambiguities while keeping them in easily understandable clauses. This standard also needs to be produced by a convenient standardization body. For STDMA this can be an issue since the standard is both within the telecommunication area and the aeronautical area. These issues are examined in more detail in this subclause. C.2 ETSI requirements for a clear standard C.2.1 The need for clear interfaces STDMA is intended to support different services and applications. We find mainly: Automatic Dependent Surveillance Broadcast (ADS-B), Differential GNSS, SMGCS, ATN. The standardization work will have to define clear interfaces between these services and applications and the STDMA communication system. First one has to collect the various services that STDMA may be foreseen to support. To the previously mentioned services one may add expected services to be supported in the future. For example, systems based on SDTMA are already being developed to support maritime and land mobile applications and hence it is possible that STDMA might also need to support these applications. Precise requirements are to be defined encompassing size of packets, generation process, requested access delay, reliability requirement, distribution mode etc. Second, these requirements are to be translated in well defined interfaces. Preferably one shall try to group the supported functions in order to rationalize the interfaces. For example, one might make use of the well established telecommunication characteristics: • connected versus connection-less function; • multipoint versus point-to-point function; TR 101 130 V1.1.1 (1997-11) 39 • periodic or asynchronous data transfers; to simplify the various interfaces. C.2.2 Clear boundaries for the system Apart from the crucial necessity to clearly define interfaces, one also needs to define exactly the system that we are aiming to standardize. This system may range from a simple VHF link operating on given frequencies within a given frequency band to a more complex system able to handle various frequencies. We may envision a system where the telecommunication system will precisely handle the coexistence of many services using a same resource. Actually we can distinguish roughly three main architectures: • Architecture 1: The first way is to define a very simple communication system capable of supporting various functions. This communication system will operate at a given frequency. This frequency may be within a given frequency window. • Architecture 2: The second way is to define in addition to the above mentioned simple telecommunication system additional tools which make it possible to manage various frequencies and to organize various functions sharing a same resource. In other words, management tools should be defined which can be used in an upper layer to integrate various functions. • Architecture 3: The third way is to build a full system which will support various applications. This way will make it possible to fully organize various functions sharing a same resource. For instance this approach will manage possible competition between functions and will be likely to guarantee quality of services to the various supported applications. This approach is distinct from the second one in the sense that integration of the various functions will be organized within the communication system itself; primitives and schemes which will handle this system will likely be integrated in the low layers : LLC, MAC layers and even may be in the physical layer. The choice between these various approaches will clearly be a committee choice. Among the points which can lead to a choice we may find: • status of existing services and applications to be supported by STDMA; • stability or possible evolution of these services and applications; • reliability required by these services and applications; • technical consideration about resource requirements of the various supported services; • level of competition for the resource of the various service. One may give simple hints to help this choice. If the service to be supported may change a lot that is in favour of architecture 1 or 2. Conversely if the requirements of the services are very stringent this suggests the use of architecture 3 since one will have a better control of the whole system. However a refined technical analysis will indicate if a precise control of the system is possible above the LLC layer. C.2.3 Clear mode of operation in the given spectrum Concerning a radio telecommunication system, one has to define the available spectrum to support this application and the acceptable out of band radiated power. Another very important aspect is the way the spectrum will be used. Will STDMA be used in an exclusive way? Will STDMA suffer from low power interferers? Will STDMA have to coexist or inter operate with another transmission system? This needs to be carefully studied during the standardization process. C.2.4 The reliability issue The reliability issue is not usually a major point for a telecommunication system based on a shared medium. However, because of the particular integrity requirements in the aeronautical field, the reliability issue is certainly to be carefully TR 101 130 V1.1.1 (1997-11) 40 studied. This issue will have to be addressed throughout all the standard. However, the following precise examples are given for illustration: The first example concerns the reservation scheme. First of all we can notice that the correct operation of STDMA requires that the reservations of slot are correctly and consistently registered. We then need this procedure to be tested to ensure that all the STDMA equipment produced by different manufacturers respect minimum rules. These rules will ensure that the system shows the required reliability. The reliability issue means also that the system needs to be able to cope with faults. Let us try to illustrate this problem by different situations. In a first situation, we will have to react to possible intrusion or jamming in the system. In this case we will have to define special operational rules. For instance, after a long jamming on a channel, we will have to restart completely the access to this channel. Moreover, the way the nodes will access it will follow special rule to avoid collisions. In a second situation, we may be in a situation where one station fails and transmits for a long period. This transmission does not correspond to real packet transmission but to a faulty operation of the system. In this case the standard should mandate a procedure to stop this faulty behaviour (a similar procedure can be found in the current aeronautical VHF voice communications system). The third situation concerns the problem of the timing synchronization. Reference [2] proposes the use of more than one source of time. Let us suppose that we have a primary and a secondary source of timing. Let us also assume (see annex D) that the primary source of time provides an absolutely synchronized timing as the secondary time source provides a relatively synchronized time source. Then the following situation may occur if the primary timing fails but non uniformly in the network. We have then nodes receiving both the primary and secondary timing sources and other nodes which will receive only the secondary timing source. There is an issue since the secondary source does meet the requirement of an absolutely synchronized timing. Therefore if the nodes do not uniformly use the secondary timing source the requirement of a relative synchronization is not met. Therefore we will need special procedure to cope with this case. The reliability issue is of course linked to the ability to ensure the correct operation of the system. This is the area of the conformance testing. It will be necessary to produce conformance testing specification attached to the functional specifications. One can notice possible standardization difficulties. For instance, there is an issue on how the final system would be tested. One may want to test the whole set (device plus antennas plus aircraft) since the proper protocol operations needs a correct position estimate and an isotropic antenna coverage (in order to fit CCI conditions on slot pre-emption). Therefore there would be a need to specify an anisotropy radiation tolerance window. C.2.5 Structure and presentation of the standard The standard needs to be written in accordance with the chosen architecture. The selected interfaces need to be carefully described. The organization of the document needs to reflect the architecture choices of the system in order to ease the comprehension and to reduce the risk of misunderstanding. Moreover for the simplicity of the presentation basic primitives need to be identified. This will reduce the size of the standard by reducing inappropriate repetitions. The standard will use also the ETSI presentation. C.3 Illustration of previous points on the current ICAO standard C.3.1 Clear interfaces We can not identify in the current draft ICAO standard clear and well defined interfaces. For instance although most of the necessary parameters of the functions are defined, the operation of the functions in case of failure is not described. For example, there is insufficient detail of the response if a function can not be provided or the response does not come in time. In addition the functions parameters are numerous and as are the number of functions described. One may think that a simplification could be useful. TR 101 130 V1.1.1 (1997-11) 41 C.3.2 The system boundary Apparently the system defined by the current draft ICAO standard [1], [3] only specifies a simple communication system. This is not in conformance with reference [2] (appendix B), where notions of channel management are introduced. Moreover the functions described in the current ICAO standard [1], [3] (SARPs) are the basis of a multi application telecommunication system. But it is not described how these different functions can coexist. For instance will they use the same channel or different channels? How will the quality of service be insured? C.3.3 Reliability procedures We do not find in the current draft any procedures related to the reliability issue. For instance, if a node suffers from a clock failure, this failure may damage the framing structure and may spoil other node transmission. The STDMA standard needs to take such situations into account by defining appropriate procedures to cover the most obvious case of failures leading to performance degradations. We also note that there is no conformance test associated to the draft standard. C.3.4 Structure and presentation of the standard The presentation of the current draft standard does not satisfy the requirement of ETSI presentation. In an ETSI standard one first describe the conditions under which the procedure is invoked. Then the procedure itself is described. Moreover to ease the understanding the notation takes into account the layer in which the procedure operates. The same procedures are described many times. The following illustration gives an example procedure in which the reservation is registered in the reservation table. Reservation information recording: This procedure is executed to record the reservation information of a received STMPDU in the local STM-entity's reservation information base upon receipt of a STMPDU. Procedure: The STMPDU is received from a neighbouring STM-entity, which is identified by the source address parameter of STC–UNITDATA indication primitive delivering the received STMPDU. The destination STM-entity of the STMPDU is identified by the destination address parameter of the STC-UNITDATA indication primitive delivering the received STMPDU. The number of the slot where the STMPDU has been received is identified by the slot number parameter of the STC-UNITDATA indication primitive delivering the STMPDU. If the destination address parameter of the STC-UNITDATA indication primitive delivering the received STMPDU is a unicast address, then a new reservation entry is recorded in the local reservation information base for a holding time tRE, where: • Rsource is set to the source address parameter of the STC-UNITDATA indication primitive; and • Rdest is set to the destination address parameter of the STC-UNITDATA indication primitive; and • Rslot is set to the slot number parameter of the STC-UNITDATA indication primitive delivering the received STMPDU plus the reservation offset value of the received STMPDU reservation offset field modulo M1. While recording this new reservation entry, an earlier reservation entry with the same Nsource and same Nslot, if it exists, is considered outdated and is replaced. If the destination address of the STC-UNITDATA indication primitive delivering the received STMPDU is not a unicast address, then a new reservation entry is recorded in the local reservation information base for a holding time tRE, where: • Rsource is set to the source address parameter of the STC-UNITDATA indication primitive; and • Rdest is set to broadcast reserved destination address value; and TR 101 130 V1.1.1 (1997-11) 42 • Rslot is set to the slot number of the STC-UNITDATA indication primitive delivering the received STMPDU plus the reservation offset value of the received STMPDU reservation offset field modulo M1. While recording this new reservation entry, an earlier reservation entry with the same Nsource and same Nslot, if it exists, is considered outdated and is replaced. TR 101 130 V1.1.1 (1997-11) 43 Annex D: Technical issues D.1 Introduction The aim of this clause is to identify the potential technical issues that should be addressed in order to: • specify the missing procedures in the current ICAO draft; • fix the parameters in the standard in order to optimize the system; • demonstrate that the system meets the system service requirement; • check that the system is technically viable and to evaluate its performance. D.2 Missing specifications D.2.1 Secondary timing and positioning protocols specification It is said in the ICAO draft standard that a node can retrieve a slot synchronization and its position from the reception of packets from the other nodes. These procedures are not described in the draft standard. Five techniques have been identified [2] to synchronize timing: • nominally, independent users are synchronized by their local GNSS equipment, providing UTC time; • synchronization from ground stations transmitting synchronization broadcast bursts; • use of low-cost atomic clocks; • synchronization from users transmitting synchronization broadcast bursts; • a fallback synchronization mode called "floating network" in which each user synchronizes from the others. Techniques 1 and 3 are related to primary timing, Technique 2 and 4 are secondary timing procedure. Some of these techniques can also be used to derive secondary position estimates. Note that position information is required to support the slot-sharing algorithms in STDMA, although these algorithms probably only require crude position estimates, perhaps to 1 nmi accuracy. Secondary position estimation may however be important in backing up the position estimates for ADS-B applications, and, for this reason, higher accuracy may be required. Notice that technique 3 does not provide a position estimate. Technique 2 will provide position estimate but in this case there should be more than 3 ground stations in range. Technique 5 is related to distributed clock synchronization which will be discussed in system enhancement subclause (see subclause D.4.2.4). Below we discuss secondary timing and positioning. We basically focus on technique 4 since technique 2 is just a subset of technique 4 restricted to ground stations. If there are three or more remote stations in range (aircraft or ground stations), the unsynchronized station can make a secondary time estimate by using those remote stations as timing and positioning beacons. Indeed the remote stations periodically broadcast their respective positions. Upon reception the unsynchronized station can compute the deltas of the propagation delays from accurate measurement of the burst starting time at its own antenna. The delta estimates together with the position indication of the remote station will provide an estimate of the secondary timing and position. In case there would be only two remote stations in range, the equation with the deltas contains an extra unknown parameter. To get rid of this unknown the unsynchronized station could use an RSSI estimate of the distance with the remote stations which will translate the measure of burst signal power at local antenna into metric distance to remote transmitter. TR 101 130 V1.1.1 (1997-11) 44 In order to avoid "tertiary" timing being derived from a secondary timed station, with a resultant further reduction in accuracy, it should be necessary that stations indicate in their position bursts the nature of its timing (secondary or primary). A study therefore needs to be carried out to address the timing and positioning recovery procedure when the local UTC source fails. A clone of the GNSS procedure based on position broadcast of remote aircraft and tracking of their burst synchronization may suffice. The main difficulty is in the definition of hardware and software timing requirements for this purpose and in providing tests to demonstrate compliance. D.2.2 Channel management The channel management is an important issue of the STDMA standard. Actually the offered bandwidth on a given channel is not sufficient to cover all the requirements that may be addressed to STDMA. Therefore it is necessary for the different functions to be able to share different frequencies. This issue is of course linked to the clear definition of the STDMA system concept. Mainly we can distinguish between two approaches to this problem. The first approach, which we call 'distributed approach', will be based on the main principle according to which the nodes will use an additional frequency when the load on the already occupied bandwidth has reached a given threshold. This approach requires that primitives providing the load status of the system be defined. This approach has the following advantages: • it is simple, the channel management is not directly tied with the access scheme; • there is no need for a master station. This approach has the following drawbacks: • the distributed selection of the additional channel may cause the problem of coherency and fault tolerance; • the primitive which will indicate if a new frequency is necessary may be complicated since the quality of service requirements are different for the various applications. For instance the ADS-B requirements are more stringent than a simple packet exchange. The second approach, which we call 'centralized approach', is based on the main principle according to a master node will control the use of the different channels. This node will have the knowledge of all the resources required by the other nodes and can therefore manage the various frequencies to satisfy the requests. This approach has the following advantage: • the master station having the knowledge of the whole system can use the frequencies very effectively; • the management related to the quality of service can be more effective; • there is no fault tolerance problem except the failure of the master station; • it is easier to handle the coexistence with a ground control mode in that case. This approach has the following drawbacks: • it is rather complicated approach, the channel management will be mixed with the access scheme; • how can we choose the master station? This station can be a ground station but in that case all the node of the network must be within range. If more than one master station is needed, these stations must react coherently; • cannot work over sea or areas without ground infrastructure. Both of these approaches have to manage the problem of jamming on a given bandwidth. This problem can be difficult if the jamming is only local and not well spread on the network. TR 101 130 V1.1.1 (1997-11) 45 The set of channel management procedures is very important. If the committee plans to include them in the draft standard then there will be the need to be study to specify numerous protocols for dynamic channel assignment and for dealing with possible erroneous behaviours of the system in case of protocol failure. D.2.3 High priority pre-emption procedure A procedure that allows a station holding a high priority burst to steal the reservation slot of lower priority belonging to another user is claimed but not yet described. Algorithms exist in case of a centralized protocol where a central agent rules the medium access. In a distributed protocol such as that used for STDMA, the problem might be less easy since the pre-emption poses problems in the case of contention between different priorities. This issue should be studied during the standardization work. D.3 System optimization D.3.1 Physical layer The following issues need to be resolved: • power limits for in band emission due to modulation and switching (in dBm versus distance to carrier frequency); • power limits for out of band emissions; • ramp-up and ramp-down timing tolerances (upper bound to be quantified, lower bound would come from in-band and out-of band power limits requirement); • receiver power threshold spacing (-87 dBm, -90 dBm, -92 dBm); • limits of frequency change during emission. D.3.2 Link layer The following issues need to be resolved: • maximum clock shift rate with respect to UTC; • retransmission parameters; • random access parameters. D.3.3 Summary Optimization of parameters in the draft ICAO standard should not require too much resource, provided the committee has the technical expertise to achieve it, or can rely on technical database and simulation tools. A number of STDMA simulation and modelling studies are currently in progress as part of the ICAO process and it is hoped that this expertise can be drawn upon to carry out the work. D.4 System enhancements D.4.1 Physical layer No significant issues were found during the review. TR 101 130 V1.1.1 (1997-11) 46 D.4.2 Link layer D.4.2.1 Superframe parameters Parameter M1 could be fixed in order to set superframe duration to exactly one minute. Doing so will relieve the ambiguity about absolute numbering of slot and the fact that each beginning of M1/60 slots coincide with the start of an UTC second. In this case it will suffice to specify that that the start of any UTC second shall coincide with a beginning of slot as it is the case in the current STDMA draft. Some clarification is therefore required of the superframe parameters. This is not expected to require significant resource to solve. D.4.2.2 Warming up procedure A system starting from scratch must wait for a superframe duration before getting any access right. This duration could be considered as being too long. Furthermore the process may not be efficient in case of many stations simultaneously starting from scratch. This could be case when the channel has been disabled after a long jamming period. A real improvement of the system should be to find a faster warm up procedure in order to increase safety. A possible way might be to specify a reservation map exchange procedure, or to reserve some slots in the channel for new aircraft to enter the system. However, it should be noted that most aircraft will enter the system during its start up procedure on the ground, and hence the warming up time may not be significant. A study needs to be carried out to investigate this issue and possibly to propose an enhancement to the standard. D.4.2.3 Slot reuse Slot reuse is when a new station can successfully transmit a burst on an already reserved slot without damaging the transmission of the previous user. The draft standard allows slot reuse but the new station needs to compute both its own signal decays and signal decay of the other user in order to compare them at the antenna of the intended receiver of the other user. Then it needs to check that its own transmission will not prevent the correct reception from the other user on its intended receiver: i.e. to let the latter to receive the signal from the other user C times greater than the signal of the new station. C is determined by the CCI conditions. When the user is transmitting in broadcast mode the condition is equivalent to guaranteeing a circle of correct reception around the transmitter. Therefore the worst case receiver will on the circle on the segment between the new station and the other user. D.4.2.3.1 Superframe multiple slot reuse The draft standard specifies only the case when there is only one alternative user on the same slot. But since the protocol allows slot reuse and requires to track multiple reservations of the same slot, the CCI condition should also take into consideration the sum of all the simultaneous signals arriving on the receiver antenna, and not only make the comparison with only one targeted user. Notice that the case of broadcast may not reduce to simple distance ratios comparison. D.4.2.3.2 RSSI based slot reuse versus position based slot reuse The draft standard requires that stations use aircraft position broadcast and assumes isotropic antenna coverage and free space propagation in order to compute signal decays. Therefore the CCI condition deals with distance ratios. An alternative procedure should be to use RSSI instead of position and free space propagation assumption. The station would maintain a RSSI table whose entry will be every neighbour stations and whose data would be the RSSI power measurement of the burst received from this neighbour. The RSSI measurement is a good indication of distance to a remote transmitter, and if the reverse light path principle applies, then it is a good estimate of the power at which the remote user would receive a signal from the host station. If there would be a way to exchange such RSSI information, then the CCI condition will only consist to add and compare the RSSI estimate on targeted receivers. For broadcast it would suffice to compare the RSSI of the burst in the current reserved slot with a simple threshold. TR 101 130 V1.1.1 (1997-11) 47 The advantage of RSSI based reuse are as follows: • it does not rely on position broadcast, since this position broadcast would rely on CCI condition it removes a potential vicious circle in reliability; • it allows to simplify the problem of testing antenna isotropic coverage (with the difficulty of testing this on the aircraft itself); • it simplifies the broadcast CCI condition; • it makes it possible to cope with the problem of possible discrepancy between free space propagation and actual propagation conditions, for example, in case of ground effects. The drawbacks are as follow: • The RSSI information table exchange may be costly (but is not needed for broadcast operation). The issue of slot re-use therefore need substantial study and validation through modelling and should include consideration of an RSSI approach. Once again, existing simulation work could be used to assist this work. D.4.2.4 Distributed software clock synchronization A precise timing is a key point of the STDMA protocol. The issue of maintaining a synchronized timing in a distributed system is a classical problem [11]. There are two kind of clock synchronization requirements: the absolute synchronization and the relative synchronization. In absolute synchronization there is an universal time, for instance UTC, that every clock must indicate, within a specified tolerance window. In relative synchronization it is only required that all clocks indicate the same time, whatever be this time, within a specified tolerance window. What is needed in STDMA is relative synchronization but with absolute rate: i.e. the clock rates are aligned with UTC rate within a tolerance window (see reference [15] for the formal definition of these two kinds of synchronization). In fact the relative synchronization needed in STDMA should be weaker, since it suffices that all clocks indicate starts of slot at the same time, regardless of the local number of this slot. An absolute synchronization requires an absolute time source. This source may be given by a broadcast system for example the GNSS or a set of ground stations having an absolute reference. The satellite system or the ground station system periodically broadcasts the absolute time within the network. Every node accurately updates its local clock provided it can infer its propagation delays. An absolute synchronization can also be achieved by independent local clock, for instance atomic clocks, if their rates are steady enough to keep the synchronization during long periods of time. In this time atomic clocks should have to be re-synchronized before each flight. A relative synchronization can be achieved with independent clocks even if their precision is not very high. This is a classical problem in the field of distributed computing systems [12]. In that case one needs to implement a synchronization algorithm which will maintain the clocks synchronized. This algorithm will be based on the exchange of synchronization packets which, to simplify all the nodes, periodically broadcast their local time. This exchange of packets is called a round. At the end of each synchronization round each node reads the clocks of all processes and then adjusts its clock value for the next round by applying a convergence function. With such algorithms, it is possible to get a precision in the synchronization which depends of the difference between the propagation delays, the time interval between two successive rounds and maximum clock rate discrepancy. One can find academic works which describe how to cope with the case where clocks fail or if packets are lost [14]. It is also well studied how the function can be chosen in order to provide the better synchronization [16], [17]. The drawback of such algorithms is that they theoretically require that the medium access be independent of the slot synchronization, which is not the case with STDMA which would need slot synchronization to communicate. A possible short-cut could be to allow STDMA to operate in degraded mode when there are clock synchronization problems. There is also an issue if one has two timing sources and if the secondary timing source is only relatively synchronized. In that case the nodes of the network need to use consistently either the primary timing source or the secondary timing source. Actually it is not obvious that the primary timing source will consistently disappear. We need to know what timing source we use in the network. Therefore, a timing mechanism that does not rely on an external source might be investigated to increase the integrity of the system. Software synchronization mechanisms provide interesting results but require a priori a communication TR 101 130 V1.1.1 (1997-11) 48 whose reliability is independent of clock synchronization. Investigations to find a possible less demanding protocol are recommended. D.4.3 Reaction to long jamming periods This is a priori a modulation issue, the modulation scheme needs to be resistant enough to cope with persistent noise. But if a persistent jamming occurs there would be a need to adjust the power threshold used for detecting idle/busy slot status. For example the threshold would rise when the minimum detect level during a long period is too high. Doing so will always allow stations to transmit anyhow, maybe with a smaller efficient range. If the jamming occurs at a high power level which is to high as to prevent correct reception of packets, then there will be the need to fall back on an alternative channel. In this case a blind channel selection protocol should be specified as part of channel management. Of course this might not be sufficient if the jammer creates white noise with large bandwidth. D.5 Viability check and performance studies D.5.1 Choice of the modulation scheme ARINC characteristic 750 is the proposed 31,5 kbps D8PSK solution for VDL modes 2 and 3. Its advantages are: • relatively high data rate for a 25kHz channel, making it a good choice for point-to-point data link systems (note that final validation of adjacent channel interference levels has not yet been satisfactorily demonstrated and hence there is some doubt about whether D8PSK data can co-exist adjacent to current voice channels); • active development in support of mode 2 and mode 3 equipment; • adoption as the ICAO approved data link. Its disadvantages are: • poor CCI performance making it less well suited to single channel broadcast applications where frequency re-use is a fundamental issue; • poor CCI also means that, even if distinct coverage cells are used, there is an increased distance between cells using the same frequency: this may become problematic in Europe given the shortage of frequencies; • it is a complex modulation scheme whose realization requires relatively sophisticated processing. This increases the cost of the equipment and may be a barrier to General Aviation users; • suspected poor linearity and thermal problems; • not proven to fall within 25kHz channel. An alternative to ARINC 750 is GFSK modulation at 19,2 kbs [7]. This is being proposed as one scheme supported by mode 4. The advantages of GFSK are: • relatively good CCI performance (in theory, validation is not yet complete although tests are underway) which makes it more suitable for applications where frequency re-use is an issue (e.g. ADS-B and to a lesser extent point-to-point); • because of its frequency re-use characteristics, it is claimed to be more spectrum efficient than D8PSK and therefore to achieve a higher data rate over a large geographical area (see references [5] and [6]); • it can be realized using simple and cheap hardware. Its disadvantages are: • lower data rate within a single cell making it less suitable for point-to-point applications if the system uses distinct coverage cells; TR 101 130 V1.1.1 (1997-11) 49 • not an approved ICAO scheme. A conclusion at the recent AMCP WGD meeting in Madrid was that D8PSK would not be a practical modulation scheme for ADS-B because of its poor CCI characteristics (the assumption here is that ADS-B would use global channels) but that it would probably be the best for point-to-point communications using distinct coverage cells. The choice of optimum scheme should therefore address the balance between the basic data rate at the physical layer and the ability of the modulation scheme to support frequency sharing. STDMA is sufficiently flexible to support two modulation types. The choice of modulation type may need further study. However for ADS-B and DGNSS (NABS) applications, it is expected that GFSK will prove the offer the best system performance. A study needs to be carried out to investigate the relative advantages and disadvantages of the possible modulation types and to make recommendations for the best scheme as a function of service provided. D.5.2 Impact on performance of CCI conditions The distinction between broadcast and unicast for slot reuse may not be worthy in terms of performance. A simpler scheme based on power measurement and adaptive would probably provide similar performance (see subclause D.4.2.3). D.5.3 Broadcast range and reliability If we refer to E-TDMA study there is a need to guarantee a reliability and a minimal range for correct reception. These requirements would be necessary in order to use STDMA as basis of ADS-B application. The E-TDMA study [10] notices that the reliability should be higher than 1-10-6 for a maximum outage duration of 30 s. A minimal interpretation of this requirement could be that at least one position broadcast should be received in any random period of 30 s with probability higher than 1-10-6. The E-TDMA suggests the following worst case en-route scenario: • a density of 570 air planes per 160 nmi radius air cylinder; • a position broadcast period of 6 s per aircraft. Using GFSK channel parameter (i.e. 75 slots per second), the position broadcast period corresponds to an average individual load of µ = 2,22 × 10 -3 per slot, considering one channel for ADS-B. For two channels this data should be divided by two. A first remark is that it is impossible to guarantee a minimal successful coverage for a broadcast burst. Indeed there is a gracefully decreasing function p(r) which provides the probability of successful reception of a broadcast burst by a random receiver at distance r. If p(r) <0,9 then it will mean that a minimum of six position broadcasts would be needed during any period of 30 s in order to achieve the 1-10-6 reliability target, assuming the broadcasts are independent (i.e. include some randomness). Since the specified number of broadcast per 30 s is five the reliable range should be set at p(r) >0,94. Function p(r) will decay as function of actual traffic load decreasing the broadcast reliability. From function p(r) one derives the average reception area σ per broadcast. In the following we present a simple model as an example of study which could be carried out with respect to this topic. Of course this simple model relies on simple assumptions. A more accurate model would need much more involvement. Simple model We assume that a receiver receives successfully a burst when the signal of the latter is greater than C times the cumulated signal of the other simultaneous reception. We assume that at each slot there is a Poisson density of transmitters of λ per unit area. If the area unit is the square nmi the E-TDMA worst case assumption [10] will lead to λ = 1,57 × 10-5. We neglect the propagation delays. Our aim is to derive an estimate of function p(r). From this quantity we can deduce the average area σ of reservation successful reception: TR 101 130 V1.1.1 (1997-11) 50 σ π = ∞ ∫ 2 0 rp r dr ( ) We investigate two propagation models: the α attenuation model and free space attenuation model with horizon. D.5.3.1 The α attenuation model In this model we assume that signal decays in r −α with distance r , with α greater than 2. Let W be the measure of the cumulated signals received by a random receiver on a random slot. Quantity W is a random variable, we define its Laplace-Stieljes transform w(t) as the average value of e tW − . Assuming that the circles around the receiver are independent: w t tx xdx ( ) exp( (exp( ) ) = − − − ∞ ∫ 2 1 0 πλ α easy calculations yield: w t t ( ) exp( ( ) ) / = − − πλ α α Γ 1 2 2 where Γ(.) is Euler's Gamma function. This expression can be used to get asymptotic estimates of the function of distribution of W. Let F(x) denotes the probability that W be greater than x, we have: F(x) ≈ − − − − λπ λ π π α α α α α x x 2 2 2 4 2 4 4 1 2 / / sin( ) ( ) ( ) Γ Γ Since p r F Cr C r ( ) ( ) / = − ≈ − 1 1 1 2 2 α α λπ we obtain the asymptotic equivalent of p(r) when r tends to 0. Figure D.1: p(r) as function of r for λ = 1,57 × 10-5, α = 2,7, C = 10, r in nmi In the figure above we display a plot of function p(r) with the load conditions of E-TDMA scenario. We have also plotted the estimated threshold for the 1-10-6 reliability per 30 s target. Notice that it provides a reliable range of 14 nmi. TR 101 130 V1.1.1 (1997-11) 51 Figure D.2: 3-D version of Figure D.1 Figure D.3: Reservation failure probability (1-p(r)) with λ = 1,57 × 10-5, α = 2,7, C = 10 The average successful area has an exact expression: σ λ = − B 1 , with B C = 2 2 2 sin( / ) / π α α α . The quantity λσ is constant and can be called the reuse factor. Notice that this quantity tends to zero when α tends to 2. TR 101 130 V1.1.1 (1997-11) 52 Figure D.4: Reuse factor as function of parameter α For α = 2,7 and λ = 1,57 × 10-5 we find σ = 6 214 sq nmi. If the individual load µ is given the average number of aircraft correctly receiving a broadcast would be B µ , for α = 2,7 this number would be 44. By pure symmetry argument this number should also be equal to the average number of aircraft detected and correctly received by a random station. Notice that the quantity is far below the worst case 570 aircraft in 160 nmi range of E-TDMA study. In case these figures were confirmed by a more involved studies, it would imply that ADS-B needs several channels in parallel, at least 11. Indeed the number of detected aircraft in presence of n parallel channels would be n B µ . The figure below display the average number of detected aircraft as a 2D function of α and individual broadcast period y in seconds. Figure D.5: Average number of detected aircraft versus α and individual broadcast period y (in seconds) Notice that the longer the broadcast period the larger is the number of detected aircraft, but this to the cost of a lower reliability. TR 101 130 V1.1.1 (1997-11) 53 D.5.3.2 The free space propagation with horizon When α = 2, equivalent to free space propagation, the above integrals diverge. Therefore in free space propagation it is needed to introduce an horizon R due to Earth roundness beyond which two aircraft are not in line in sight and therefore cannot receive signal from each other. Assuming an altitude of 10 km we obtain an horizon of 400 nmi for the aircraft and 200 nmi for the ground stations. The Laplace-Stieljes now reads: w t tx xdx R ( ) exp( (exp( ) ) = − − − ∫ 2 1 2 0 πλ which can be rewritten in: w t t t R ( ) exp( ( )) = −πλ ψ 2 with ψ ( ) ( ) z e dx x x z = − − +∞ ∫1 2 . The interesting case is when the horizon is large. When R is large, to be more precise when the horizon load λπR2is large we have the following estimate: F x x R O x R ( ) log( ) ( log ( )) ≈ − + λπ λπ λπ λπ 2 2 2 2 1 which is valid when x increases significantly above the logarithm of the horizon load. Therefore when r tends to zero we have: p r Cr r C R ( ) log( ) ≈ − − 1 1 2 2 2 λπ λπ λπ . The data for worst case scenario in E-TDMA study provide a horizon load estimate of horizon load estimate at 8. Figure D.6: The function p r( ) versus r (in nmi), with horizon load equal to 8, C=10, free space propagation TR 101 130 V1.1.1 (1997-11) 54 Figure D.7: 3-D version of Figure D.6 Figure D.8: Reservation failure probability (1-p(r)) versus in nmi, with horizon load equal to 8, C=10, free space propagation Remark the above estimate provides a reliable range of 10 nmi. The average successful reception area σ has an asymptotic expression which is: σ π λ λπ λπ ≈ + 1 1 2 2 2 C R O R log ( ) (log ( )) the utilization factor can be derived as well. Notice that the reuse factor does not converge to a constant but decreases when the horizon load increases. TR 101 130 V1.1.1 (1997-11) 55 Figure D.9: The reuse factor as a function of horizon load, free space propagation We notice that the reuse factor with the horizon load of 8 is around 0,015. With the same horizon load, which corresponds to the worst case scenario of E-TDMA we find an average reception area of 977 sq nmi. The average number of detected aircraft would be 7. This number is small and is probably inaccurate and should be taken carefully. Figure D.10: Average number of detected aircraft versus horizon load H and individual broadcast period y (in seconds) D.5.3.3 Conclusion of the model The simple model is independent of the access scheme because the captured effect is the cumulative effect of transmissions from distant aircraft using a single frequency. The results would be the same with any another access scheme. The model needs more detailed development and we must be cautious about the precise result. In particular, there is a need to take account of the distribution of traffic since a significant volume of traffic is low level and hence below the horizon for distant users. However one do sees that there is a dimensioning problem to be investigated in order to determine the number of channels required to support the expected applications. For instance, the model illustrated the kind of analysis which would help the determination of the number of needed parallel channel for ADS-B. As a matter of fact one can notice that n channels in parallel also divides the horizon load by n in the expression of the reuse factor. More refined models will be necessary to dimension and to design the system for proper operation. TR 101 130 V1.1.1 (1997-11) 56 D.5.4 The effect and management of remote and hidden terminals When two stations are at distance greater than 300 km (160 nmi), the propagation delays exceed 1 ms and the bursts are mis-synchronized in such a way that they occupy two consecutive slots on the receiver side. Therefore a protocol malfunction will occur that should be analysed since the load due to remote stations might be not negligible. Maybe D8PSK/GFSK guard times versus received power thresholds could be optimized. For a system using global channels and providing ADS-B information it is necessary to co-ordinate the access to the channel so as to prevent some users being unable to receive position information on other users. This "hidden terminal" problem arises when two users broadcast in the same slot with the result that other users receive a transmission from only one user, or in the worst extreme, neither user. In a ground controlled system, it is relatively straightforward to define a co-ordination mechanism to ensure that all users that can be heard by a ground station can also hear each other. However, this does not ensure that mobiles can receive other mobiles that are not within reach of the same ground station and hence there is some risk to the integrity of the ADS-B air-air application. Note that mobiles are better placed to make decisions on channel usage since they receive transmissions from the greatest number of transmitters, many of which are not received by the ground station because of ground shielding (which should be investigated in details). STDMA uses a self-organizing access scheme that combines access control by mobiles with the option of ground control where practical. The scheme has been demonstrated at low traffic levels but will probably require further refinement to deal with higher traffic levels and to achieve high integrity performance. For example the possible discrepancies between airborne reservation map and ground station reservation map (leading to hidden nodes cases) could be removed if one allows aircraft to forward their own reservation map to the ground station. Alternatively, an aircraft at higher altitude than the hidden transmitters could notify those of the garbling situation and order some to change slots. The hidden node problem depends also of the required Signal over Noise ratio. The larger is the latter the higher will the collision rates due to hidden nodes. In this perspective D8PSK would perform less than GFSK. TR 101 130 V1.1.1 (1997-11) 57 History Document history V1.1.1 November 1997 Publication ISBN 2-7437-1798-X Dépôt légal : Novembre 1997 |
aadef7d867a9a9c9597d9cf09763d938 | 101 105 | 1 Scope | The present document describes the requirements (at a stage 0 level) of the Fraud Information Gathering System (FIGS). FIGS provides the means for the HPLMN to monitor a defined set of subscriber activities. The aim is to enable service providers/network operators to use FIGS, and service limitation controls such as Operator Determined Barring (ODB) and Immediate Service Termination (IST), to limit their financial exposure to large unpaid bills produced on subscriber accounts whilst the subscriber is roaming outside their HPLMN. HPLMNs may also choose to collect information on subscriber activities whilst their subscribers are within the HPLMN. |
aadef7d867a9a9c9597d9cf09763d938 | 101 105 | 2 References | The following documents contain provisions which, through reference in this text, constitute provisions of the present document. • References are either specific (identified by date of publication, edition number, version number, etc.) or non-specific. • For a specific reference, subsequent revisions do not apply. • For a non-specific reference, the latest version applies. • For this Release 1999 document, references to GSM documents are for Release 1999 versions (version 8.x.y). [1] GSM 01.04: "Digital cellular telecommunications system (Phase 2+); Abbreviations and acronyms". [2] GSM 02.33: "Digital cellular telecommunications system (Phase 2+); Lawful Interception - stage 1". |
aadef7d867a9a9c9597d9cf09763d938 | 101 105 | 3 Definitions and abbreviations | |
aadef7d867a9a9c9597d9cf09763d938 | 101 105 | 3.1 Definitions | For the purposes of the present document the following terms and definitions apply: monitored activities: subscriber activities that must be reported to the HPLMN. These can be call related events (e.g. call-set-up, call termination) or the invocation of call related and call independent supplementary services (e.g. Call Hold, Call Waiting, Call Transfer, Call Forwarding, Unstructured Supplementary Service Data) Home Network: home PLMN including non-GSM elements such as the Fraud Detection System (FDS), customer service systems and billing |
aadef7d867a9a9c9597d9cf09763d938 | 101 105 | 3.2 Abbreviations | Abbreviations used in the present document are listed in GSM 01.04 and the following apply: FDS Fraud Detection System This is not necessarily an automatic system but may be one that requires human intervention. FIGS Fraud Information Gathering System IST Immediate Service Termination ETSI ETSI TR 101 105 V8.0.0 (2001-05) 6 (GSM 01.31 version 8.0.0 Release 1999) |
aadef7d867a9a9c9597d9cf09763d938 | 101 105 | 4 Fraud Information Gathering System overview | A number of proposals have been suggested for a Subscriber Supervisory System (SSS) for which specifications were produced from May 1995 through to December 1996. Following joint review between SMG1 and SMG10, it was agreed that the system should be re-specified to take account of network operator and manufacturer needs for a Fraud Information Gathering System (FIGS). The present document provides an outline of such a system. The present document describes a method by which the Home Network can be provided with data on the activities of its subscribers in a VPLMN. The Home Network can make inferences about what the subscriber is doing and then take decisions on what the subscriber should be allowed to do. The present document does not address any Fraud Detection systems or the intelligence that is used to advise the HPLMN on the controls to be applied to a subscriber. Figure 1 shows the flow of messages between the HPLMN and the VPLMN and between the HPLMN and the FDS. Fraud Detection System HPLMN FIGS Set FIGS Data VPLMN 1 VPLMN 3 FIGS Set FIGS Data VPLMN 2 Figure 1: Flow of messages between the HPLMN and the VPLMN and between the HPLMN and the FDS |
aadef7d867a9a9c9597d9cf09763d938 | 101 105 | 5 The need for fraud detection systems and controls | |
aadef7d867a9a9c9597d9cf09763d938 | 101 105 | 5.1 Outline of present situation | Modern telecommunications networks, particularly mobile networks provide the potential for fraudsters to make use of telecommunication services (Voice, Data, Fax etc.) without the intent to pay. A number of different scenarios are exploited and it is up to the network operator or service provider to detect misuse where it occurs and to stop it at the earliest possible opportunity. The scale of frauds can be many thousand of ECU per day on a single account when International or Premium rate numbers are called. The most common types of fraud that effect networks like GSM are related to the ability to sell calls at below market price using stolen air-time/equipment where the user of the equipment does not intend to pay the network operator or service provider. Fraudulent subscribers often avoid payment by obtaining a handset and a subscription to a GSM network by fraudulently giving details and justifications to the network operators/service provider. If there are not good controls within the network the subscriber can make a large volume of calls to expensive destinations and accumulate a large bill. ETSI ETSI TR 101 105 V8.0.0 (2001-05) 7 (GSM 01.31 version 8.0.0 Release 1999) Roaming, in co-ordination with advanced services such as call transfer and multi-party calls, complicates the issue further, requiring control of the customer within the VPLMN. Many simultaneous calls can be set up and large bills accumulated in a short time. At present no system exists within the GSM network architecture for speedily transferring information on subscriber activity from the VPLMN to the HPLMN. In the future, SIMs may roam to non-GSM networks, further broadening the area over which control is required. It is recognised that if FIGS is implemented in non-GSM networks that suitable inter-working units will be required to translate commands and information. |
aadef7d867a9a9c9597d9cf09763d938 | 101 105 | 5.2 General Principles | The PLMN network should be able to supply relevant information to the HPLMN network so it can make a decision on whether to terminate a call or to change the Operator Determined Barring (ODB) configuration for the specific subscriber. This decision will be carried out by the HPLMN or service provider. It is recognised that there is a limit to the type and volume of information that can be transferred between the VPLMN and the HPLMN. Therefore the requirement for the system is that distilled and standardised information must be supplied between the VPLMN and HPLMN. |
aadef7d867a9a9c9597d9cf09763d938 | 101 105 | 5.3 Capabilities | The following minimum capabilities are required. See figure 1. Within the Home Network: - to mark a subscriber, defined by the IMSI or MSISDN, as being under FIGS control ("FIG Set"); - to receive from the VPLMN the data described below; - to remove the monitoring of a subscriber's activities ("FIGS Unset"). Within the VPLMN: - to transmit to the HPLMN information (FIGS Data): - at the start of a call; - at the end of a call; - during a call' for long calls or at the mid-call invocation of supplementary services. |
aadef7d867a9a9c9597d9cf09763d938 | 101 105 | 5.4 Service conditions | The following service conditions shall apply: - FIGS shall not modify the VPLMN's service; - FIGS should not alter any standard GSM functionality seen by the customer or effect the service quality; - If the VPLMN network does not have the resources to support a FIGS Set command it shall respond accordingly to the HPLMN. |
aadef7d867a9a9c9597d9cf09763d938 | 101 105 | 5.5 Information Delivery Time | The need for up to date information is a critical part of any fraud information system. The sooner data is transferred to the HPLMN, the sooner fraud can be stopped. Therefore the proscribed information shall be transferred from the VPLMN to the HPLMN within two minutes of the occurrence of a FIGS-monitored event. The information shall preferably be transferred from the VPLMN to the HPLMN over existing communication links (e.g. SS7 signalling links). ETSI ETSI TR 101 105 V8.0.0 (2001-05) 8 (GSM 01.31 version 8.0.0 Release 1999) |
aadef7d867a9a9c9597d9cf09763d938 | 101 105 | 5.6 Subscriber Data Volumes | If the support of FIGS is causing overload within the VPLMN the FIGS system shall not permit the marking of new subscribers. The VPLMN should therefore handle up to a realistic limit any requests for marking of subscribers and be able to support the associated data transfer. The setting of this limit is outside the scope of the present document. Each VPLMN should limit the number of subscribers that each HPLMN may request to be monitored using FIGS. Otherwise an HPLMN may take more than its "fair share" of the FIGS processing capability of a VPLMN. A mechanism shall be required whereby a VPLMN can charge an HPLMN for the bulk data transfer made to that HPLMN. |
aadef7d867a9a9c9597d9cf09763d938 | 101 105 | 6 Interface between HPLMN and FDS | The interface between the home network and the network's fraud detection and processing systems shall be through a specific interface. This will be used to present information to the fraud detection systems. The contents of messages sent on this interface shall be specified but not the transfer mechanism. This is in line with the approach used for the X-interface as specified in GSM 02.33. The FDS will indicate to the HPLMN subscribers that should be subject to FIGS monitoring. This information will update the HPLMN HLR. Information, as listed in clause 5.3 gathered from the VPLMN will be transferred to the FDS system. Following processing of this information, the FDS system can take no action or can advise the home network to do one of the following: a) update ODB categories; b) instigate an Immediate Service Termination (IST); c) mark the subscriber as not being required to be monitored under FIGS. |
aadef7d867a9a9c9597d9cf09763d938 | 101 105 | 7 Security of the system | It is expected that there will be a need for authentication, data integrity and confidentiality of the commands and data transferred between PLMNs. These issues are for study under other work items within the SMG10 work programme. ETSI ETSI TR 101 105 V8.0.0 (2001-05) 9 (GSM 01.31 version 8.0.0 Release 1999) Annex A (informative): Status of GSM 01.31 Status of Technical Report GSM 01.31: stage 0 of FIGS Date Version Remarks No Phase 1 version June 1997 1.0.0 To SMG#22 for information October 1997 2.0.0 To SMG#23 for approval October 1997 5.0.0 TS approved by SMG#23 March 1998 7.0.0 The report was converted to version 7.0.0 because the work item is related to Release 98. Version 5.x.y was withdrawn (SMG#25) June 1998 7.0.1 CR 01.31-A001 (Editorial) approved by SMG#26 April 2000 8.0.0 The report was converted to version 8.0.0 because the work item is related to Release 1999 Text and figures: WinWord 7.0 Stylesheet: etsiw_70.dot Rapporteur: Tim Wright (Vodafone) ETSI ETSI TR 101 105 V8.0.0 (2001-05) 10 (GSM 01.31 version 8.0.0 Release 1999) History Document history V8.0.0 May 2001 Publication |
9ec8c7518d16e2fcc1cd4499a28ba3c3 | 101 122 | 1 Scope | The present document is to analyse the impact of Service Provider Portability on Geographic and non-geographic numbers and number formats used at the Network Termination Point (NTP) and also at the Point Of Interconnection (POI) between networks in a multi-vendor environment. Routeing requirements are analysed and numbers and number formats are derived as a consequence. |
9ec8c7518d16e2fcc1cd4499a28ba3c3 | 101 122 | 2 References | References may be made to: a) specific versions of publications (identified by date of publication, edition number, version number, etc.), in which case, subsequent revisions to the referenced document do not apply; or b) all versions up to and including the identified version (identified by "up to and including" before the version identity); or c) all versions subsequent to and including the identified version (identified by "onwards" following the version identity); or d) publications without mention of a specific version, in which case the latest version applies. A non-specific reference to an ETS shall also be taken to refer to later versions published as an EN with the same number. [1] ITU-T Recommendation E.164: "International Telecommunication Numbering plan". [2] TR NA 010063: "High level Description of Number portability". [3] TR NA 010064: "High level Architecture and solutions to support Number Portability". |
9ec8c7518d16e2fcc1cd4499a28ba3c3 | 101 122 | 3 Definitions and abbreviations | |
9ec8c7518d16e2fcc1cd4499a28ba3c3 | 101 122 | 3.1 Definitions | addressed entity: Any entity identified by an address in the routeing process (e.g. called party, serving exchange, point of interconnection, IN-element - depending on the routeing method). directory number: The number that is dialled by the users to reach the called customer (potentially with prefix and/or with suffix). routeing number: A specific number that is added and used by the networks to route the call. The Routeing Number conveys information usable by the network. If the digits dialled by the user match the digits of a routeing number, the dialled digits should not be interpreted as a routeing number. ported number: A number that has been subject to number portability. routeing information: Information needed to complete the call. It consists of Routeing Number (RN), Directory Number (DN) or RN + DN. service provider: An entity that offers services to users involving the use of network resources. The "Service Provider" is understood in the present document in a generic way and may have different status according to the service provided. For example, "Service Provider" refers to a local loop operator in the case of Geographic Numbers, or to a mobile operator in the case of Mobile Numbers, or to a service operator / reseller in the case of Service Numbers. donor network: The initial Network where a number was allocated by the Numbering Plan Administrator before ever being ported. TR 101 122 V1.1.1 (1997-11) 6 recipient network: The Network where a number is located after being ported. serving network: The network that determines whether a number has been ported, and, if so, provides an appropriate routeing number. This functionality may be distributed. transit network: A network between two networks, e.g. . the recipient network and the donor network. donor exchange: The initial Exchange where a number was located before ever being ported. recipient exchange: The new Exchange where a number is located after being ported. serving exchange: A Serving Exchange) is, within this document, an exchange within a Serving network (SN) that makes a data base (Exchange internal or external) access to retrieve a Routeing Number for a call to a portable number. database query function: The function whereby a database is accessed in order to ascertain whether a number is ported, and if it is, a Routeing Number is obtained that may be used to route the call to a destination. The database could form part of an IN implementation, could be embedded within the switch, or could be some form of other off-switch database. national numbering plan: A national Numbering Plan is a scheme that structures the numbers used and the numbers space available in a country. network termination point: The point where a call is delivered. point of interconnection: An access point between two networks. |
9ec8c7518d16e2fcc1cd4499a28ba3c3 | 101 122 | 3.2 Abbreviations | For the purposes of the present document the following abbreviations apply: CC Country Code (E.164) CgPN Calling Party Number COLP COnnected Line identification Presentation DN Directory Number NNS National Numbering Scheme NP Number Portability NPA Numbering Plan Administrator N(S)N National Significant Number (E.164) NTP Network Termination Point POI Point Of Interconnection RN Routeing Number SP Service Provider |
9ec8c7518d16e2fcc1cd4499a28ba3c3 | 101 122 | 4 General assumptions and guidelines | |
9ec8c7518d16e2fcc1cd4499a28ba3c3 | 101 122 | 4.1 General guide-lines | The main purpose of the present is to describe types of numbers/addresses to be used by the callers and also by network operators in order to set-up calls to ported numbers. The main routeing problem to solve when Number Portability (NP) is involved is to be able to route the call to the correct exchange in the correct network and select the access line of the subscriber with the ported number in the recipient exchange |
9ec8c7518d16e2fcc1cd4499a28ba3c3 | 101 122 | 4.1.1 General Assumptions | The present document deals with call set-up and analyses what types of numbers/addresses should be used at various reference points in the overall call path between calling and called party in order to establish calls. It also generally considers addressing and numbering in the context of number portability at SCCP level. TR 101 122 V1.1.1 (1997-11) 7 |
9ec8c7518d16e2fcc1cd4499a28ba3c3 | 101 122 | 4.1.2 Guide-lines for number types specification | 1) All numbers identifying a network termination point are ITU-T Recommendation E.164 [1] numbers (i.e. comply with ITU-T Recommendation E.164 [1] requirements), this may not be the case for numbers only carried within a network or transferred from one network to another (e.g. these numbers may use non-decimal digits or may have other formats). 2) A number which can be dialled by a caller is a disable number. This category includes numbers normally used to set-up calls and also any number format which can be received by a local exchange and trigger a routeing process. Signalling protocols may provide an indicator to distinguish disable numbers from others. 3) If the digits dialled by the user match the digits of a routeing number, the dialled digits should not be interpreted as a routeing number. |
9ec8c7518d16e2fcc1cd4499a28ba3c3 | 101 122 | 4.2 Assumptions made as regards routeing | The following general routeing scheme is assumed as the routeing model for calls routed to a ported customer. incoming call Serving Network Serving Exchange Recipient Network Recipient Exchange 2nd step of the Routeing process, based on routeing number 1st step of the Routeing process, based on dialled digits customer with ported number Note : The serving Network Transit Network (optional) may be the originating Network and/or the donor or a transit network Figure 1: Conceptual framework for incoming calls 1) The caller sets-up the call by dialling the DN as usual. The DN is enough to initiate the routeing process. Furthermore, number portability, by definition implies that the callers should continue to dial the same DN and nothing more to set up a call to a ported customer. TR 101 122 V1.1.1 (1997-11) 8 2) The routeing process is split into 2 consecutive main steps: a) normal routeing based on DN towards a serving exchange. As a 1st step in the routeing process, the originating network routes the call up to a serving exchange clearly identified by the analysis of a certain number of leading digits of the DN. b) routeing to customer's interface based on number(s) obtained by this serving exchange. It should be noted that this step might be subdivided into sub-steps (e.g. the serving exchange could provide information to route to a database - within the recipient network or accessed by the recipient network - which provides subsequent routeing information identifying the recipient exchange, information used for a following sub-step in the routeing process). 3) If only the recipient network is identified, then it is the responsibility of the recipient network to obtain the subsequent RN to terminate the call at the recipient exchange. 4) In any case the internal routeing process in the recipient exchange shall unambiguously terminate at the called customer's interface. 5) If a number is ported subsequently from Service Provider #1 to Service Provider #2, then to Service Provider #3, etc., this will change the Routeing Number but not change the routeing principles. |
9ec8c7518d16e2fcc1cd4499a28ba3c3 | 101 122 | 4.3 Assumptions made as regards the numbers | 1) Incoming calls to a ported number are set-up by callers dialling DN in either the local or the national format. For incoming international calls, the caller abroad dials +CC National Significant Number (N(S)N), the DN format handled in the national network is the national format N(S)N. 2) Outgoing calls set-up by a customer with a ported number benefit from the CLI function i.e. the calling number is optionally forwarded up to the called party and may be used by the called party or by any involved network for various purposes (e.g. by a transit network for carrier selection billing). The calling number indicated should therefore be the DN of this customer in any case. 3) the structure and the format of the RN should be, in a given country, independent of the network architecture to support NP. The structure and format of RN are unique in one country if passed between networks. 4.4 Assumptions made as regards the numbering scheme management Some objectives may be assumed as regards the management of the national numbering plan: 1) minimize the impact on the National Numbering Scheme (NNS) (e.g. minimize the amount of additional numbers needed by the networks); 2) provide a clear distinction between the non-directory numbers and directory numbers of the NNS , to help their management; 3) if a number is ported several times (i.e. from Service Provider #1 to Service Provider #2, then to Service Provider #3, etc.), the subsequent portability should not increase the amount of routeing numbers needed beyond what is needed for the initial portability. |
9ec8c7518d16e2fcc1cd4499a28ba3c3 | 101 122 | 4.5 Other considerations | Ported numbers are assigned a routeing number in their new Service Provider's domain. In a general case there is a many-to-one mapping between DN and RN, but in special cases there may be a one-to-one relationship. TR 101 122 V1.1.1 (1997-11) 9 |
9ec8c7518d16e2fcc1cd4499a28ba3c3 | 101 122 | 5 Addressable entities for routeing purposes. | Entities which need to be addressed by a RN in one or more routeing solutions are identified in this clause. According to the structure of the routeing number, one or a combination of several of the following entities should be addressable. Recipient Network: in this option, the routeing number identifies the network where the customer is now located. Therefore the routeing process will need an additional information (i.e. DN) to be completed. Point Of Interconnection (POI): in this option, the routeing number identifies an interface to the next network in the routeing process. Therefore the routeing process will need an additional information (i.e. DN) to be completed. Recipient exchange: in this option, the routeing number identifies the exchange the customer is now located. Therefore the routeing process within the recipient exchange will need an additional information (i.e. DN) to be completed. Network Termination Point (NTP): in this option, the routeing number identifies the Subscriber/Access line/service. The ported customer identified by the RN is unique. Therefore the routeing process, in terms of Number Portability, can be completed without any additional information. Combined entities: In this option, we may use RN to identify any combination of the above entities. 6 Types of addresses and numbers — within networks and across network boundaries With service provider portability it is no longer possible to use the Directory Number, dialled by the calling party, to route the call to the customer. An additional information, the RN, is needed to be able to route the call. The Routeing Information may have one of the following : - concatenated address (subclause 6.1); - separated address (subclause 6.2); - partly separated address (subclause 6.3); - only RN, i.e. plain network address, suppressed ITU-T Recommendation E.164 [1] number (subclause 6.4); - only DN, i.e. plain ITU-T Recommendation E.164 [1] number (subclause 6.5). It shall be taken into account that in case of the concatenated and partly separated addressing schemes limitations can be present on the maximum numbers of digits being supported by the signalling system and the exchanges in the different networks involved. |
9ec8c7518d16e2fcc1cd4499a28ba3c3 | 101 122 | 6.1 Concatenated address |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.