hash stringlengths 32 32 | doc_id stringlengths 7 13 | section stringlengths 3 121 | content stringlengths 0 2.2M |
|---|---|---|---|
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.4.4.1 Transmitter Side Coupler Configuration for Transfer Function Measurements | TX side Delta sym. P-N EP NE PN IN slide switch lever position P E N CM S1 S3 S4 S6 S7 50 50 LVDN TX side Delta sym. N-E EP NE PN IN slide switch lever position P E N CM S1 S3 S4 S6 S7 50 50 LVDN TX side Delta sym. E-P EP NE PN IN slide switch lever position P E N CM S1 S3 S4 S6 S7 50 50 LVDN Figure 6: Coupler Configuration: Feed Differentially |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.4.4.1.1 Slide Switch Positions | Table 8: Switch Positions of Transfer Function Measurements Feed PN Feed NE Feed EP P (S4) E (S5) N (S6) CM (S7) off off off on E-P (S1) P-N (S2) N-E (S3) on on on P (S4) E (S5) N (S6) CM (S7) off Off off on E-P (S1) P-N (S2) N-E (S3) on On on P (S4) E (S5) N (S6) CM (S7) off off off on E-P (S1) P-N (S2) N-E (S3) on on on |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.4.4.2 Receiver Side Coupler Configuration for Transfer Function Measurements | PN EP NE CM S1 S3 S4 S6 S7 OUT P E N RX side Star + CM slide switch lever position LVDN Slide Switch Positions P (S4) E (S5) N (S6) CM (S7) on on on on E-P (S1) P-N (S2) N-E (S3) off off off When P, E and N are measured CM must not be terminated. The cable at CM port must be disconnected. When CM is measured P, E and N must be terminated with 50 Ohm. Figure 7: Coupler Configuration: Receive from Single Conductor and CM ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 17 |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.4.5 Conducting Channel Transfer (S21) Measurements | If the equipment is set-up and the network analyzer calibrated as described above field measurements can be conducted in private residential units. A protocol-sheet is prepared in clause A.2 for each measurement site, to be filled out during field tests. S21 have to be measured with every combination of feeding (NE, PN, EP, APN, PNE) and receiving (P, N, E, CM). To protect the NWA from damage, the coupler should be connected to the outlet before the coaxial wire is connected and when removing the coupler from the outlet, the coaxial cable should be disconnected first. |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.5 Reflection (S11) Measurements | |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.5.1 Measurement Principle | The LVDN is a network with an undefined complex characteristic impedance. The often measured absolute value of the input impedance has little practical significance. Adding a short piece of mains cable may change the results considerably. Thus STF 410 measured the reflection loss S11 at the 'Delta' terminals of the Universal couplers instead of the impedance. 50 Ohm E-P N-E P-N balun TRANSMIT interface (Delta) T1 T2 T3 50 Ohm 50 Ohm E P N to LVDN NWA S11 200 Ohm balun balun S1 on S2 off S3 off 200 Ohm 200 Ohm of Universal coupler Figure 8: Principle for DM Impedance (S11) measurement via the Baluns of the Universal Couplers (Example: P_N) ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 18 |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.5.2 Set-Up | At S11, reflection measurements signals are fed and received at one and the same coupler. Coupler A Tx+Rx LVDN Network Analyzer Isolation Filter for DM + CM (e.g. LISN) Power Supply Figure 9: Set-up S11 Measurements: Feed and Receive Differentially The file name convention of reflection (S11) measurements is: Pt_Fa_Rb_xx.xx.CSV where: • 't' is the number of the transmitting plug. • 'Fa' is the port where signals are fed differentially: EP, PN, NE, EPNT, PNNT, NENT, APN, PNE and CM. When feeding 'NT' is used the two other ports at the delta coupler should not be terminated. This allows a comparison with SISO PLC. The individual feeding styles are introduced in clause 6.1.5 in [i.6]. • 'Rb' is the port where signals are received differentially: EP, PN or NE. For reflection (S11) measurements the identical feeding and receiving port is used. 'a' is identical to 'b'. • 'xx.xx' is the timing distance to the rising LCZC at the Tx coupler in ms when the sweep was recorded. If the NWA trigger was not in sync with LCZC 'xx.xx' should not be applied. E.g. if the filename is P2_PN_PN.csv the reflection was recorded at Plug number 2 differentially between Phase and Neutral. All reflection (S11) measurements should be saved in the 'S11' folder of the STF410 data repository. |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.5.3 Calibration of NWA | The NWA has to be calibrated with a full 1-port reflection calibration at the end of the coaxial wire. As calibration kits 'short', 'open' (do not connect anything) and 'broadband load' (50 Ohm termination) has to be used. In the measurements recorded here the MIMO PLT couplers [i.5] are considered to be part of the PLT channel. S11 is a complex value which is a function of the load impedance and of the characteristic impedance of the measurement system. In our case the measurement system consists of the network analyzer, which has a characteristic impedance of 50 Ω and the 1:2 balun inside the Universal coupler, which transforms the 50 Ω to 200 Ω. S11 on the 50 Ω side is identical to S11 on the 200 Ω side, except for a phase shift due to the length of the transmission lines inside the balun. ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 19 The real and the imaginary parts of S11 are recorded. The absolute value is │S11│ = sqrt((real(S11)^2 + imag(S11)^2)) and the phase angle is φ = atan(imag(S11)/ real(S11)) For engineering purposes the absolute value │S11│is sufficient in most cases. It allows us to calculate the load impedance depending on the line length. ZDMmax = Zo (1 + │r│) / (1 - │r│) with Zo = 200 Ω and │r│ = 10 (- │S11│/20) Knowing φ one may calculate ZDM for each frequency when considering the line length of the balun which is about x = 0,3 m. ZDM = Zo (1 + │r│* e j(φ + βx) )/ (1 - │r│* e j(φ + βx)) where β = ω / υ υ: speed of propagation in the balun. |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.5.4 Functional Test Before Starting Reflection (S11) Measurements | From time to time, one should perform a │S11│check, before starting reflection measurements with the LVDN, by connecting the mains plug of the coupler to the test pad. The characteristic DM-impedance of the pad is 80 Ω. Ideally │S11│ is - 7,4 dB. Because of loss and impedance mismatch in the coupler and of the Schuko plug the value is somewhat frequency dependent. Impedance P-N EP NE PN IN slide switch lever position 19.2 dB Test pad P E N CM S1 S3 S4 S6 S7 open slide switch lever position 19.2 dB Test pad EP NE PN IN P E N CM S1 S3 S4 S6 S7 Impedance N-E open slide switch lever position 19.2 dB Test pad EP NE PN IN P E N CM S1 S3 S4 S6 S7 Impedance E-P open Figure 10: Coupler Configuration: Feed Differentially |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.5.4.1 Slide Switch Positions | Table 9: Switch Positions of Functional Test for Reflection Measurements Feed & Receive PN Feed & Receive NE Feed & Receive EP P (S4) E (S5) N (S6) CM (S7) off off off on E-P (S1) P-N (S2) N-E (S3) off on off P (S4) E (S5) N (S6) CM (S7) off off off on E-P (S1) P-N (S2) N-E (S3) off off on P (S4) E (S5) N (S6) CM (S7) off Off off on E-P (S1) P-N (S2) N-E (S3) on Off off |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.5.4.2 Typical Return Loss for Inputs P-N; N-E and E-P | Table 10: Return Loss of Functional Test for Reflection Measurements 10 30 80 MHz │S11│ 7,2 to 7,5 7,7 to 8,6 7,5 to 10,5 dB ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 20 |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.5.5 Conducting Reflection (S11) Measurements | If the equipment is set-up and the network analyzer calibrated as described above, field measurements can be conducted in private residences. A protocol-sheet is prepared in clause A.2 for each measurement site to be completed during field tests. All combinations of feeding and receiving (NE, PN, EP, NENT, PNNT, EPNT, APN, PNE, CM) are recorded. Any combination where feeding and receiving did not take place on the same port (line conversion reflections) was not measured by STF410. To protect the NWA from damage the coupler should be connected to the outlet before the coaxial wire is connected and when removing the coupler from the outlet the coaxial cable should be disconnected first. |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.6 Set-Up for Noise Measurements | |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.6.1 Set-Up | The full setup for noise measurement is given in figure 11. DSO PROBE Filters AMP BOARD PWR SUPPLY Figure 11: Noise Measurement Setup The MIMO coupler is used in receiver mode in the star configuration. In order to receive signals on 4 ports, including the CM signal, the P, N and E switches need to be closed and the CM switch left open ('on' position). The filters are used in 4 different configurations called Band 1 to Band 4 as given in table 11. Table 11: Filter Configuration for each Frequency Band Band Filter configuration Comment Band 1 HPF-002 + LPF-100 2-100 MHz band Band 2 HPF-002 + LPF-100 + SBF-FM 2-100 MHz band with FM notch Band 3 4HPF-025 + LPF-200 30-100 MHz band Band 4 HPF-025 + LPF-100 + SBF-FM 30-100 MHz band with FM notch ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 21 0 5 10 15 x 10 7 -120 -100 -80 -60 -40 -20 0 20 40 Frequency 20* log|S21| AMP_28DB.CSV BG_88_108MHZ.CSV HP_25MHZ.CSV HP_2_150MHZ.CSV LP_100MHZ.CSV Figure 12: Frequency Response at 0 < f < 150 MHz of Filters and Amplifier used After the filter stage, each signal is connected to the DSO using this convention: • Port 1: P signal • Port 2: E signal • Port 3: N signal • Port 4: CM signal At the beginning of the measurement, the settings of the DSO should be set as follows: Vertical settings: the vertical gain of each port is set to the minimum, that is, the observation scale is set to the maximum. In the case of the LeCroy WaveRunner 64Xi VL, the maximum observation scale is 1,00 V/div. Horizontal settings: the sampling frequency of the DSO is set to 500 MS/s. The observation duration is set to 20 ms (i.e. 2 ms/div on most DSOs). In this configuration, the memory requirement for each port is 10 MS. Termination: the ports of the scope should be terminated internally with 50 Ohm. Amplifiers are optionally used when the received signal is too weak. Thus, the measurement process for each band should be as follows. 1) Connect the coupler to the outlet (see important notes about equipment attachment below). 2) Connect the coupler outputs to the filters inputs and the filters outputs to the DSO ports 3) Observe the noise measurements on the DSO and, for each port individually, increase the vertical gain in order to span the signal across the full observation scale. Make sure that there is no clipping of the received signal on any port. 4) If the noise level for all 4 ports is < 10 mV (measured before the amplifiers), attach optional amplifier board and repeat step 2 (see important notes about equipment attachment below). 5) Record noise waveform. 6) Detach optional amplifier board (see important note about equipment attachment below). 7) Detach DSO and filters from coupler. ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 22 8) Detach coupler from outlet. Important notes about equipment attachment: • Connecting and disconnecting the coupler from the outlet creates impulsive noise that can damage the equipment. Thus, connecting the coupler to the outlet should be the first step in the equipment attachment process. Similarly, disconnecting the coupler from the outlet should be the last step in the equipment detachment process. • Powering up the amplifier board without a load on the amplifier output can damage the amplifiers. So powering up the amplifier should be the last step in the equipment attachment process. Similarly, powering down the amplifier from the outlet should be the first step in the equipment detachment process. • Operating conditions of each Mini-circuit ZFL-500LNB+ power amplifier are 15 V and 60 mA. The amplifier board (consisting of 4 amps) should be fed with a stabilized power supply delivering 15 V and 240 mA. Depending on the type of power supply, it might be necessary to check these values using a multi-meter. In order to synchronize all measurements with the LCZC, the DSO trigger must be set up and the mains line signal must be used at zero crossing (with a positive slope) as a trigger. To ensure that all 4 paths are recorded at the same time, a 'single trigger' shot has to be recorded, before saving the data. The file name convention for noise measurements is: Paa_BDb_Gcc_d_yyyy.eee where: • 'aa': Two digit Receive Plug (outlet) number • 'b': Filtered band number (1 to 4) • 'cc': Two digit amplifier gain in dB. Usually 'cc' is 28 when amplifiers are present and 'cc' equals 00 if the amplifier is not used • 'd': Signal at Rx Star- coupler: P, E, N or CM • 'yyyy': if the recording was done in a special environment, this may be noted here. E.g. 'yyyy' may be 'all appliances_on' or 'noise_from_PC_power_supply'. If several noise shots are recorded using one and the same setting, an index of 'yyyy' can be used • 'eee': File extension: 'trc' for LeCroy DSO, 'wfm' or 'isf' for Tektronix DSO E.g. if the filename is P04_BD2_G28_E.trc, receiving was done at Plug number 4 at connector E (Protective Earth). The filtered band is BD2, i.e. 2-100 MHz with FM notch. An amplifier with 28 dB gain was used. Please verify that the y-axis settings are included in the record. If not, please include the voltage per division settings in the filename. |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7 General Equipment List | |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7.1 Coaxial Cables | The coaxial cables used to conduct the measurements must enable results of a dynamic range of up to 120 dB. Therefore double shielded cables like RG214 are required. |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7.2 Network Analyzer | The following NWA are used by the measurement teams. ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 23 |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7.2.1 Agilent E5071B | The team in Germany used an Agilent E5071B. Figure 13: Agilent E5071B Table 12: Technical Properties of Agilent E5071B Property Value Comment Type E5071B ENA Manufacturer Agilent Output power +12 dBm in the frequency domain; into 50 Ω Out- / Input impedance 50 Ω Frequency range 300 kHz to 8,5 GHz Max Dynamic Range 125 dB |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7.2.2 Agilent E5071C | The teams in Belgium and France, Lannion used a Agilent E5071C ENA Network Analyzer. Figure 14: Agilent E5071C Table 13: Technical Properties of Agilent E5071C Property Value Comment Type E5071C ENA Manufacturer Agilent Output power +10 dBm in the frequency domain; into 50 Ω Out- / Input impedance 50 Ω Frequency range 9 kHz to 6,5 GHz Max Dynamic Range 123 dB ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 24 |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7.2.3 Rohde & Schwarz ZVB4 | The team in Spain used a Rohde & Schwarz ZVB4 Network Analyzer. Figure 15: Rohde & Schwarz ZVB4 Table 14: Technical Properties of R&S ZVB4 Property Value Comment Type ZVB4 Manufacturer Rohde & Schwarz Output power -40 dBm to + 13 dBm Out / Input Impedance 50 Ω Frequency Range 300 kHz to 4 GHz Max Dynamic Range > 123 dB at 10 Hz IF bandwidth Number of ports 4 Number of measurement points 1 to 60 001 |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7.2.4 CM Choke Absorber between Coupler and NWA | To ensure that coupler and the counterpoise provide a set-up for reproducible measurements, CM signals from the coupler need to be isolated from NWA. In the measurement configuration shown in figure 16, the counterpoise is decoupled by chokes with absorbent ferrites. The choke, or chokes in the case of multi-channel measurements, should be placed close to the RX coupler. Z DM Z CM NWA coupl. TX coupl RX LVDN earth counter poise I CM absorber choke TX RX NWA RX TX DM- CM- filter if connected close to TX and RX Figure 16: Improved Measurement Configuration ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 25 Figure 17 shows a possible design with a toroidal ferrite core and a RG-316 or similar cable. Figure 17: Choke Absorber CM Impedance of Choke Absorber (recorded choke data with 4 turns) Table 15: Impedance of CM Choke 3 10 30 60 80 100 MHz │Z│ 255 660 920 530 370 250 Ω φ 78 38 2 - 30 - 60 - 50 degree |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7.3 Digital Sampling Oscilloscope | The following DSO are used by the measurement teams. |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7.3.1 Tektronix DPO4104 | The team in southern Germany used Tek DPO4104. http://www.tek.com/products/oscilloscopes/mso4000/http://www.tek.com/products/oscilloscopes/mso4000/ Figure 18: Tek DPO4104 ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 26 Table 16: Technical Properties of Tektronix DPO4104 Property Value Comment Type DPO4104 Manufacturer Tektronix No of Channels 4 Plus external trigger Input impedance 50 Ω Resolution 8 bit Additionally 'High Resolution' option averaging values acquired by oversampling Max Bandwidth 1 GHz Memory depth 10 M samples for each of the 4 channels |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7.3.2 Tektronix DPO7254 | The team in Belgium / North-Western Germany /United Kingdom used a DPO 7254. See http://www.tek.com/products/oscilloscopes/dpo7000/. Figure 19: Tektronix DPO725 Table 17: Technical Properties of Tektronix DPO7254 Property Value Comment Type DPO7254 Manufacturer Tektronix No of Channels 4 Plus external trigger Input impedance 50 Ω Resolution 8 bits (> 11 bits with averaging) Max Bandwidth 2,5 GHz Memory depth 10 GS/s for each of the 4 channels |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7.3.3 LeCroy WaveRunner 64Xi VL | The team in France, Lannion used a LeCroy WaveRunner 64Xi VL. http://www.lecroy.com/Oscilloscope/OscilloscopeModel.aspx?modelid=1939&capid=102&mid=504 ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 27 Figure 20: LeCroy WaveRunner 64Xi VL Table 18: Technical Properties of LeCroy WaveRunner 64XiVL Property Value Comment Type WaveRunner 64Xi VL Manufacturer LeCroy No of Channels 4 Plus external trigger Input impedance 50 Ω Resolution 8 bit Max Bandwidth 600 MHz Memory depth 12,5 Msamples for each of the 4 channels |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7.3.4 Agilent DSO8104A | The team in Spain used a Agilent DSO8104A. Figure 21: Agilent DSO8104A Table 19: Technical Properties of Agilent DSO8104A Property Value Comment Type DSO8104A Manufacturer Agilent Number of channels 4 Input impedance 50 Ω / 1 MΩ Resolution 8 bit Max Bandwidth 1 GHz Memory depth 8 Mpts / 4 Mpts 2 channels / 4 channels Sample rate 4 Gsa/s ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 28 |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7.4 Amplifiers for Noise Measurements | Four Mini-circuit ZFL-500LNB+ amplifiers were used by all teams conducting noise measurements. Figure 22: Array of 3 amplifiers with Power Supply Table 20: Technical Properties of Mini-circuit Amps Property Value Comment Type Wideband low noise Manufacturer Mini-circuit Max Output level ~7,5 dBm Max Gain ~30 dB The frequency transfer function of the amplifiers is depicted in figure 12. The amplifier output power was checked against varying input power levels in order to evaluate the amplifier 3 dB compression point. Observation results are given in figure 23. According to the measurements, the amplifier reaches its 3 dB compression point at an input power of -17 dBm. ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 29 Figure 23: Evaluation of the Amplifier 3 dB Compression Point |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7.5 LISN or Filter to Isolate Measurement Devices from Mains | Figure 24: AMN Table 21: Technical properties of AMN Property Value Comment Type ESH3-Z5 Manufacturer Rohde & Schwarz PE connection 50 μH "on" An additional filter must be inserted in sequence into the LISN in order to isolate the PE wire, and is described in the following clause. ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 30 |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7.6 Mains Filter | Test instruments (namely the network analyzer) connected to the mains section being tested, constitute an additional load during channel and transfer measurements and may cause measurement errors. The instruments should be connected to another mains section, e.g. in a neighboring apartment, via a mains extension cable when possible. Otherwise, the MIMO mains filter described herein can be used to minimize the influence. Even with a filter, mains outlets immediately beside feeding and receiving points should be avoided. The filter should be inserted between measurement equipment and the LISN. This filter was produced several times by STF410 and distributed to each measurement team. |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7.6.1 Schematic Diagram | NWA mains CM choke > 10 uH P N PE 22 // 22 uH // 1,2 k 22 // 22 uH // 1,2 k 3 x 10 nF max 6 A P N PE 820 k 820 k Figure 25: Schematic of Mains Filter |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7.6.2 Typical Impedances of Decoupling Components | |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7.6.2.1 R / L Combinations - Mains Side | Table 22: Impedance R/L Circuit on Mains Side of STF410 Mains Filter MHz 1,59 3 10 30 60 80 100 │Z│ 79 135 450 1 100 650 440 340 Ω φ 58 58 64 - 4 - 50 - 60 - 60 degree |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7.6.2.2 Common Mode Choke - Instrument (NWA) Side (4 turns) | Table 23: Impedance of CM Choke in STF410 Mains Filter MHz 1,59 3 10 30 60 80 100 │Z│ 110 240 610 850 580 400 310 Ω φ 89 82 42 - 16 - 51 - 62 - 64 degree ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 31 |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7.6.3 Images of Mains Filter | Figure 26: Mains Filter with Closed Cover Figure 27: Filter Inside View |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 6.7.7 Ground Plane | A huge ground plane is necessary to achieve a low impedance or high capacity connection to ground. The ground plane is especially important when trying to replicate the CM signals we received. The size of the ground plane is sufficient when human touch no longer influences measurement results. ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 32 Figure 28: Ground Plane with PLC Coupler Connected Tightly |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 7 Statistical Evaluation of Results | |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 7.1 Channel Transfer Function | S21 of MIMO PLC channels was recorded at 34 locations. In total 4 588 frequency sweeps of S21 measurements were conducted. Each sweep consists of 1 601 measurement values in frequency domain totalling 7 345 388 values used in the statistical evaluations below. The figures shown below represent the cumulative probability of measuring such an attenuation. Usually a stretched S-style line from the bottom left to top right corner is depicted in the graph. These figures may be read in the following way: • Lower left corner, the 0 % or 'definitely never' point: The attenuation was never larger than 100 dB in figure 29. • Median or 50 % point: every 2nd measured value provides higher or lesser attenuation than roughly 53 dB in figure 29. • 90 % point at yellow line in figure 29: 90 % of all measured values provide higher attenuation than 38 dB when feeding style APN is selected. • Top left corner, 100 % or 'always' point: the minimum recorded value in figure 29 is 5 dB attenuation. ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 33 -100 -90 -80 -70 -60 -50 -40 -30 -20 -10 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Cumulative Probability of Attenuation Attenuation in dB, independent of Frequency (1 - 100MHz) or Link all S21 records _APN_ _PNE_ _EP_ _NE_ _PN_ Figure 29: Cumulative Probability of Attenuation (S21) for All Feeding Styles All S21 measurements - independent of location, country or frequency – are separated into individual feeding possibilities: PN, EP NE, APN and PNE (figure 29). A zoom into the high attenuated records (figure 30) show that APN and PNE provide better PLT coverage. They are less attenuated than other feeding styles. The 2 wires used for traditional PLT modems (P-N) show the worst coverage in this case. NOTE: The unused ports at the coupler were terminated with 50 Ω in these measurements. The classical SISO style modems operate by sending and receiving signals symmetrically via P and N. Unused feeding combination in SISO, are either open or unterminated. Termination of unused ports, when identical energy is fed into the couplers, theoretically requires 1,96 dB of signal energy from mains wires compared to an unterminated coupler where no loss takes place. -90 -85 -80 -75 -70 -65 -60 0 0.05 0.1 0.15 0.2 0.25 Cumulative Probability of Attenuation Attenuation in dB, independent of Frequency (1 - 100MHz) or Link all S21 records _APN_ _PNE_ _EP_ _NE_ _PN_ Figure 30: Cumulative Probability of Attenuation (S21) for All Feeding Styles, Close-up of High Attenuated Records ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 34 Results for the possibility of receiving, identical analytical process as above, are displayed in figure 31 and figure 32. Figure 32 is a close-up of the high attenuated values. These statistics are derived from identical measurements already presented in figure 29, but sorted by individual receiving styles. Of the receiving possibilities displayed, CM reception provides the best coverage. -100 -90 -80 -70 -60 -50 -40 -30 -20 -10 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Cumulative Probability of Attenuation Attenuation in dB, independent of Frequency ( 1 - 100MHz) or Link all S21 records _P. _N. _E. _CM. Figure 31: Cumulative Probability of Attenuation (S21) of All Receiving Styles Figure 32 provides a closer look at the higher attenuated values of figure 31. It is clearly visible that CM reception provides better coverage for PLT modems. However, a low impedance connection to ground is essential for CM reception. Only a huge ground plane can support low frequencies. An HD-TV may be the only consumer electronic device, used in a private home, equipped with an adequate ground plane. At frequencies above 30 MHz the required size of the ground plane becomes smaller and which enables the use of many devices. NOTE: The coupler itself is considered to be part of the channel in these statistics. [i.5] in clause 7.1.6 shows that CM reception with the STF410 coupler increases attenuation by 3 dB to 4 dB compared to reception on P, N or E. If someone is interested in the channel attenuation without the influence of the STF 410 coupler, the values from [i.5] might be considered causing the CM line to move to the right. The gap between the CM line and the others becomes larger at higher attenuated values. ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 35 -90 -85 -80 -75 -70 -65 -60 0 0.05 0.1 0.15 0.2 0.25 Cumulative Probability of Attenuation Attenuation in dB, independent of Frequency ( 1 - 100MHz) or Link all S21 records _P. _N. _E. _CM. Figure 32: Cumulative Probability of Attenuation (S21) of All Receiving Styles, Close-up of Higher Attenuated Records The median values at each frequency, for all feeding and receiving styles are shown in figure 33. A constant trend from 5 MHz to 100 MHz can be seen in higher frequencies being more attenuated than lower frequencies. This trend seems to be independent of feeding or receiving style. This slope is 0,2 dB/MHz. 0 1 2 3 4 5 6 7 8 9 10 x 10 7 -65 -60 -55 -50 -45 -40 -35 -30 Frequency Attenuation (dB) All feedings vs. frequency _APN_ _PNE_ _EP_ _NE_ _PN_ _P. _N. _E. _CM. Figure 33: Median Value of Attenuation (S21) Depending on Frequency and Feeding or Receiving style ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 36 Figure 32 shows that the CM provides less attenuation in higher attenuated channels. Figure 34 shows a graph of the top 10 % of attenuated values depending on frequency. The phenomena of the CM having less attenuation than the differential mode signals or single ended lines (P, N or E) exists at all frequencies. 0 1 2 3 4 5 6 7 8 9 10 x 10 7 -85 -80 -75 -70 -65 -60 -55 -50 Frequency Attenuation (dB) All receivings vs. frequency _P. _N. _E. _CM. Figure 34: Statistical 10 % Value of Attenuation (S21) at Higher Attenuated Records Depending on Frequency and Receiving Style Figure 35 through figure 43 show the minimum, 20 %, median, 80 % and maximum attenuation (S21) of each feeding or receiving style depending on frequency. ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 37 0 1 2 3 4 5 6 7 8 9 10 x 10 7 -100 -90 -80 -70 -60 -50 -40 -30 -20 -10 0 Frequency _APN_ Attenuation (dB) min Att 20% Att med Att 80% Att max Att Figure 35: Min., 20 %, Median, 80 % and Max. Attenuation (S21) found per Frequency in APN Feeding Style 0 1 2 3 4 5 6 7 8 9 10 x 10 7 -100 -90 -80 -70 -60 -50 -40 -30 -20 -10 0 Frequency _PNE_ Attenuation (dB) min Att 20% Att med Att 80% Att max Att Figure 36: Min., 20 %, Median, 80 % and Max. Attenuation (S21) found per Frequency in PNE Feeding Style ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 38 0 1 2 3 4 5 6 7 8 9 10 x 10 7 -100 -90 -80 -70 -60 -50 -40 -30 -20 -10 0 Frequency _EP_ Attenuation (dB) min Att 20% Att med Att 80% Att max Att Figure 37: Min., 20 %, Median, 80 % and Max. Attenuation (S21) found per Frequency in EP Feeding Style 0 1 2 3 4 5 6 7 8 9 10 x 10 7 -100 -90 -80 -70 -60 -50 -40 -30 -20 -10 0 Frequency _NE_ Attenuation (dB) min Att 20% Att med Att 80% Att max Att Figure 38: Min., 20 %, Median, 80 % and Max. Attenuation (S21) found per Frequency in NE Feeding Style ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 39 0 1 2 3 4 5 6 7 8 9 10 x 10 7 -100 -90 -80 -70 -60 -50 -40 -30 -20 -10 0 Frequency _PN_ Attenuation (dB) min Att 20% Att med Att 80% Att max Att Figure 39: Min., 20 %, Median, 80 % and Max. Attenuation (S21) found per Frequency in PN Feeding Style 0 1 2 3 4 5 6 7 8 9 10 x 10 7 -100 -90 -80 -70 -60 -50 -40 -30 -20 -10 0 Frequency _P. Attenuation (dB) min Att 20% Att med Att 80% Att max Att Figure 40: Min., 20 %, Median, 80 % and Max. Attenuation (S21) found per Frequency when Receiving in P Style ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 40 0 1 2 3 4 5 6 7 8 9 10 x 10 7 -100 -90 -80 -70 -60 -50 -40 -30 -20 -10 0 Frequency _N. Attenuation (dB) min Att 20% Att med Att 80% Att max Att Figure 41: Min., 20 %, Median, 80 % and Max. Attenuation (S21) found per Frequency when Receiving in N Style 0 1 2 3 4 5 6 7 8 9 10 x 10 7 -100 -90 -80 -70 -60 -50 -40 -30 -20 -10 0 Frequency _E. Attenuation (dB) min Att 20% Att med Att 80% Att max Att Figure 42: Min., 20 %, Median, 80 % and Max. Attenuation (S21) found per Frequency when Receiving in E Style ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 41 0 1 2 3 4 5 6 7 8 9 10 x 10 7 -100 -90 -80 -70 -60 -50 -40 -30 -20 -10 0 Frequency _CM. Attenuation (dB) min Att 20% Att med Att 80% Att max Att Figure 43: Min., 20 %, Median, 80 % and Max. Attenuation (S21) found per Frequency when Receiving in CM Style Table 24 presents median values of attenuation at each location. The number of sweeps at each location gives an indication of the size and maturity of the statistics. Table 25 shows the median for each country of all measurements taken. The high attenuation recorded in German buildings can be explained by the 3-phase installations there. ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 42 Table 24: Median S21 Value, Size and No. of Sweeps at each Location, Independent of Feeding or Receiving Style Location Country Median S21 in dB Size of flat or building in m² No of sweeps recorded at this location Duerrbachstr Germany -57,57 115 120 ImGeiger Germany -53,91 130 120 Nauheimerstr Germany -53,23 95 220 Rothaldenweg Germany -58,16 122 180 Schlossbergstr Germany -64,02 140 60 VickiBaumWeg Germany -64,87 150 120 Boenen Germany -56,98 144 Calicanto Spain -52,63 200 120 Paiporta Spain -48,15 96 120 Valencia_46012 Spain -47,21 78 120 Valencia_46015 Spain -51,61 87 120 Xirivella Spain -51,73 110 72 Bourg-La-Reine France -41,12 85 120 Spidcom-Lab France -62,33 180 60 DeGrandRyStrasse Belgium -55,96 130 144 Heidhoehe Belgium -56,93 160 144 Huette Belgium -54,11 110 144 InDenSiepen Belgium -54,61 110 144 Simarstrasse Belgium -57,64 150 144 AlsdorferStrasse Germany -52,65 110 144 Eichelhaeherweg Germany -61,61 85 144 Schlossstrasse Germany -69,04 120 144 SchurzelterWinkel Germany -61,36 150 144 Wasserkall Germany -63,82 144 ColchesterDrive United Kingdom -60,13 145 144 WilliamsRoad United Kingdom -45,05 97 144 WindmillAvenue United Kingdom -48,24 170 144 WindsorClose United Kingdom -35,16 75 144 Devolo-Lab Germany -64,12 250 24 Cavan France -36,25 66 120 Guingamp France -45,22 135 252 RueBunuel France -35,36 120 120 RueDepasse France -54,05 120 180 Trebeurdun France -37,31 56 180 ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 43 50 100 150 200 250 -70 -65 -60 -55 -50 -45 -40 -35 -30 Size of location in m² Median S21 in dB Figure 44: Size of Location under Measurement vs. Median S21 Figure 44 displays the relationship between the size of the location under measurement and the attenuation between two outlets. Assuming a linear relationship between these two parameters a fitting line can be drafted into figure 44. The formula of this line is S21 [dB] = -0,1240 dB/m² * size [m²] -37,5546 dB Table 25: S21 Median Value and No. of Sweeps for each Country, Independent of Feeding or Receiving Style Location Country Median S21 in dB No of sweeps recorded Germany -59,06 1 708 Spain -49,93 552 France -43,42 1 032 Belgium -55,72 720 UK -47,07 576 All locations -52,69 4 588 ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 44 |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 7.2 Reflection (S11) Measurements | Reflection measurements were conducted in 33 locations in Germany, Spain, France, Belgium and the United Kingdom. In total 565 frequency sweeps have been recorded with 1 601 points in frequency domain each. This results in a statistical compilation of 904 565 values. -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Cumulative Probability of |S11| |S11| in dB, independent of Frequency (1 - 100MHz) or Link D:\svn\meas\FD\0_100MHz\STF410 _APN_ _CM_ _EP-NET_ _EPNT_ _E_ _NE-EPT_ _NENT_ _N_ _PNE_ _PNNT_ _P_ _EP_ _NE_ _PN_ Figure 45: CDF of Magnitude of S11 from all Feeding or Receiving Styles Independent of Location or Frequency Figure 45 shows an overview of the probability of measuring a reflection parameter. Indoor powerline networks show weak impedance conditions. It is difficult to implement impedance matching couplers, due to the time, frequency and location dependent characteristics which influence the coupler's feeding or receiving properties. If the S11 parameter is less (more negative) than -6 dB more than half of the feed signals are reflected back to the coupler and the connected outlet. This is the case for more than 60 % of all S11 measurements conducted using delta- or the T-feeding style. Reflection parameters of a star-style probe, the single ended lines E, N and P as well as the CM seem to provide better impedance matching than differential couplings. The T-Style coupler shows less impedance matching than the alternatives. The CDF is presented after conversion of the S11 parameter to impedance Z using Z0 = 200 Ω in figure 46. ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 45 0 50 100 150 200 250 300 350 400 450 500 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Cumulative Probability of Impedance Z Z in Ohm, independent of Frequency (1 - 100MHz) or Link D:\svn\meas\FD\0_100MHz\STF410 _APN_ _CM_ _EP-NET_ _EPNT_ _E_ _NE-EPT_ _NENT_ _N_ _PNE_ _PNNT_ _P_ _EP_ _NE_ _PN_ Figure 46: CDF of Impedance Z of all Feeding or Receiving Styles Independent of Location or Frequency Table 26 shows median impedances. Table 26: Median Impedance for each Feeding/Receiving style Feeding / receiving style Impedance in Ω all locations Impedance in Ω all locations except UK Impedance in Ω locations in UK only APN 84,24 CM 91,29 EP-NET 97,75 EPNT 90,02 E 190,41 NE-EPT 105,12 NENT 84,79 N 190,30 PNE 86,23 PNNT 84,03 P 185,65 EP 88,24 89,11 81,96 NE 86,21 87,31 78,31 PN 85,34 86,36 77,14 ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 46 The impedances of PN, EP and NE at typical locations are shown per frequency in figure 47. It is interesting to see that the variance of the impedance values becomes smaller in higher frequencies. 0 1 2 3 4 5 6 7 8 9 10 x 10 7 0 50 100 150 200 250 Frequency Impedance in Ohm all EP, NE and PN impedances in Rothaldenweg P41_EP_EP.CSV P41_NE_NE.CSV P41_PN_PN.CSV P42_EP_EP.CSV P42_NE_NE.CSV P42_PN_PN.CSV P43_EP_EP.CSV P43_NE_NE.CSV P43_PN_PN.CSV P44_EP_EP.CSV P44_NE_NE.CSV P44_PN_PN.CSV P51_EP_EP.CSV P51_NE_NE.CSV P51_PN_PN.CSV P52_EP_EP.CSV P52_NE_NE.CSV P52_PN_PN.CSV P53_EP_EP.CSV P53_NE_NE.CSV P53_PN_PN.CSV Figure 47: Impedance Z of a Typical Location The mains installation in the UK uses a ring wire on each housing level. The outlets are daisy chained along the ring. As a result each outlet is connected to two set of wires, each going in a different direction in the room. Electrical installations in the rest of Europe follow a combination of star (at the fuse cabinet) and tree style (branches into rooms, outlets, light switches, etc.). It may be interesting to see S11 and input impedance of UK installations compared to the values recorded on the continent. Table 26 also includes the median impedances of UK installations with all measurements recorded outside UK. Figure 48 to figure 51 show a comparison of S11 and Impedance Z between the UK and the continental measurements. As expected the UK installations show lower impedance but not half the value than found on the continent. ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 47 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Cumulative Probability of |S11| |S11| in dB, independent of Frequency (1 - 100MHz) or Link UK locations _EP_ _NE_ _PN_ Figure 48: CDF of S11 Recorded in UK -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Cumulative Probability of |S11| |S11| in dB, independent of Frequency (1 - 100MHz) or Link All locations without UK _P_ _EP_ _NE_ Figure 49: CDF of S11 Recorded on the European Continent ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 48 0 50 100 150 200 250 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Cumulative Probability of Impedance Z Z in Ohm, independent of Frequency (1 - 100MHz) or Link UK locations _EP_ _NE_ _PN_ Figure 50: CDF if Impedance Z recorded in UK 0 50 100 150 200 250 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Cumulative Probability of Impedance Z Z in Ohm, independent of Frequency (1 - 100MHz) or Link All locations without UK _P_ _EP_ _NE_ Figure 51: CDF of Impedance Z Recorded on the European Continent ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 49 Figure 52 to figure 60 show the frequency dependence of S11 parameters. The full set of all feeding combinations within S11 parameters was not conducted at all locations. The frequency dependent behavior of PN, NE, EP, PNE, APN, PNNT, NENT, EPNT and CM can be analyzed in the figures below. The STF 410 coupler matches the impedance of the mains best around 50 MHz at all differential mode couplings. 0 1 2 3 4 5 6 7 8 9 10 x 10 7 -40 -35 -30 -25 -20 -15 -10 -5 0 Frequency _PN_ S11 in dB max Att 80% Att med Att 20% Att min Att Figure 52: Probability of Measuring S11 at PN Dependent on Frequency (Statistics based on 139 Sweeps) 0 1 2 3 4 5 6 7 8 9 10 x 10 7 -40 -35 -30 -25 -20 -15 -10 -5 0 Frequency _NE_ S11 in dB max Att 80% Att med Att 20% Att min Att Figure 53: Probability of Measuring S11 at NE Dependent on Frequency (Statistic based on 139 Sweeps) ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 50 0 1 2 3 4 5 6 7 8 9 10 x 10 7 -40 -35 -30 -25 -20 -15 -10 -5 0 Frequency _EP_ S11 in dB max Att 80% Att med Att 20% Att min Att Figure 54: Probability of Measuring S11 at EP Dependent on Frequency (Statistic based on 138 Sweeps) 0 1 2 3 4 5 6 7 8 9 10 x 10 7 -40 -35 -30 -25 -20 -15 -10 -5 0 Frequency _PNE_ S11 in dB max Att 80% Att med Att 20% Att min Att Figure 55: Probability of Measuring S11 at PNE Dependent on Frequency (Statistic based on 28 Sweeps) ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 51 0 1 2 3 4 5 6 7 8 9 10 x 10 7 -40 -35 -30 -25 -20 -15 -10 -5 0 Frequency _APN_ S11 in dB max Att 80% Att med Att 20% Att min Att Figure 56: Probability of Measuring S11 at APN Dependent on Frequency (Statistic based on 28 Sweeps) 0 1 2 3 4 5 6 7 8 9 10 x 10 7 -40 -35 -30 -25 -20 -15 -10 -5 0 Frequency _PNNT_ S11 in dB max Att 80% Att med Att 20% Att min Att Figure 57: Probability of Measuring S11 at PNNT Dependent on Frequency (Statistic based on 18 Sweeps) ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 52 0 1 2 3 4 5 6 7 8 9 10 x 10 7 -40 -35 -30 -25 -20 -15 -10 -5 0 Frequency _NENT_ S11 in dB max Att 80% Att med Att 20% Att min Att Figure 58: Probability of Measuring S11 at NENT Dependent on Frequency (Statistic based on 18 Sweeps) 0 1 2 3 4 5 6 7 8 9 10 x 10 7 -40 -35 -30 -25 -20 -15 -10 -5 0 Frequency _EPNT_ S11 in dB max Att 80% Att med Att 20% Att min Att Figure 59: Probability of Measuring S11 at EPNT Dependent on Frequency (Statistic based on 18 Sweeps) ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 53 0 1 2 3 4 5 6 7 8 9 10 x 10 7 -40 -35 -30 -25 -20 -15 -10 -5 0 Frequency _CM_ S11 in dB max Att 80% Att med Att 20% Att min Att Figure 60: Probability of Measuring S11 at CM Dependent on Frequency (Statistic based on 48 Sweeps) |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 7.3 Noise | |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 7.3.1 Frequency Domain Noise Statistics | Noise measurements of MIMO PLC channels were recorded at 31 locations, in Germany, Spain, France, Belgium and the United Kingdom. Measurements consist of P, E, N and CM signals, each recorded using 4 different bands (see clause 6.6). In total, 2 420 time domain records of 20 ms duration each were collected. Some configurations (same location, same outlet) were recorded several times. E.g. more than one successive line cycle was recorded. The statistics presented below consider only one measurement per configuration, which leads to a statistical set of 1 928 measurements. From the time domain measurements, the noise Power Spectral Density (PSD) was computed using Welch's averaged modified periodogram method of spectral estimation. This method leads to 1 639 noise samples over a frequency band of 0 MHz to 100 MHz. Hence, the total number of noise samples is 3 159 992. For all noise samples recorded with the help of an amplifier, the amplifier gain was removed from the raw DSO measurement by post-processing. For a small number of samples (6 %), we noticed that the amplifier was not delivering full amplification, due to non-linear amplifier behavior with respect to input power (see clause 6.7.4). These samples were corrected accordingly, using a lower amplification factor. The use of the Welch method was parameterized to lead to a resolution bandwidth of 122 kHz in the frequency domain. In the figures below the graphs are converted to a PSD level in dBm/Hz to simplify comparisons of resolution bandwidths like 9 kHz or 120 kHz that are common in the world of EMC. ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 54 0 10 20 30 40 50 60 70 80 90 100 -170 -150 -130 -110 -90 -70 Frequency (MHz) Level (Resolution of 122.0 kHz) converted to PSD in dBm/Hz Band 1 max Noise 80% Noise med Noise 20% Noise min Noise Figure 61: Min., 20 %, Median, 80 % and Max. PSD Noise found per Frequency on Band 1 Figure 61 presents the statistics of the PSD noise measured for all outlets, in all locations and all Rx ports. The received Noise PSD varies between -160 dBm/Hz and -80 dBm/Hz. The presence of Short Wave (SW) broadcast frequencies is clearly noticeable at 6 MHz, 7,3 MHz, 9,5 MHz, 12 MHz, 13,5 MHz, 15,5 MHz ,17,5 MHz and 21 MHz (Meter Bands of 49 m, 41 m, 31 m, 25 m, 22 m, 19 m, 16 m and 13 m respectively). Similarly, above 88 MHz, the spectrum is occupied by FM broadcasting frequencies. It should be noted that the minimum noise level may be influenced by DSO capabilities. In particular, the quantization noise increases when measuring large voltage signals. For this reason, measurements were also taken over reduced bands. Thus, Band 2 covers the 2 MHz to 80 MHz band, Band 3 covers the 30 MHz to 100 MHz band, and Band 4 covers the 30 MHz to 80 MHz band. It is thus possible to define a composite band, selecting for each frequency the narrowest available band, to limit the signal level and thus reduce the quantization noise. ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 55 0 10 20 30 40 50 60 70 80 90 100 -170 -160 -150 -140 -130 -120 -110 -100 -90 -80 -70 Frequency (MHz) Composite band Level (Resolution of 122.0 kHz) converted to PSD in dBm/Hz Band 2 Band 4 Band 3 max Noise 80% Noise med Noise 20% Noise min Noise Figure 62: Min., 20 %, Median, 80 % and Max. PSD Noise found per Frequency over the Composite Band Figure 62 presents the same statistics as figure 61, over the composite band. When SW and FM signals are rejected, in band 4 for instance, the minimum noise PSD decreases down to -168 dBm/Hz. Using the composite band instead of Band 1 mainly affects the lowest statistics: for the median, 80 % and maximum curves, the values observed in the composite band are similar to the ones observed in Band 1. Figure 63 presents the cumulative probability of the noise PSD for all outlets, in all locations, and all Rx ports, independently of the frequency. Here again, one can observe the noise level recorded for Band 1 is somewhat larger than for the composite band, especially for the lower percentiles. This is due to the higher quantization noise of the DSO when the observation band is wider. Figure 64 gives the cumulative probability of the noise PSD, for each Rx port separately. One can clearly observe that the noise PSD recorded on the CM port is stronger than on other Rx ports, by about 5 dB. This can be due to the higher potential of the CM to receive external noise from radiating sources. In this statistical study, the coupler itself is considered to be part of the channel. From the STF410 coupler validation in [i.5], clause 7.1.6, it appears that CM reception leads to an increase in attenuation by 3 dB to 4 dB compared to receiving on P, N or E. Hence, the CM noise curve would further shift to the right with respect to the P, N or E curves when using a coupler with similar reception properties for all Rx ports. Comparing the noise PSD statistics for the P, N and E ports, it appears, that the E port is slightly more affected by noise than the P or N ports (by 1 dB) when it comes to the largest noise values (90 % percentile). For the lowest noise values (10 % percentile), the P port is slightly less affected by noise than the E or N ports (by 3 dB). ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 56 -170 -160 -150 -140 -130 -120 -110 -100 -90 -80 -70 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Cumulative Probability of Noise values All measurements, independent of Frequency ( 1 - 100MHz) or Outlet Level (Resolution of 122.0 kHz) converted to PSD in dBm/Hz Band 1 Composite band Figure 63: Cumulative Probability of PSD Noise of all Rx Ports -170 -160 -150 -140 -130 -120 -110 -100 -90 -80 -70 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Cumulative Probability of Noise values Composite band, independent of Frequency ( 1 - 100MHz) or Outlet Level (Resolution of 122.0 kHz) converted to PSD in dBm/Hz P E N CM Figure 64: Cumulative Probability of PSD Noise for each Rx Port ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 57 Figure 65 presents the median noise PSD measured on Band 1 by frequency for all outlets and all locations, separated by Rx port. Here again, we observe that the CM port is more sensitive to noise than the other ports. The gap between the CM port and other ports is larger for low frequencies below 40 MHz. For high frequencies up to the FM band, one can distinguish the P, N, E and CM ports, which are increasingly affected by noise in this order. 0 10 20 30 40 50 60 70 80 90 100 -160 -155 -150 -145 -140 -135 -130 -125 -120 -115 -110 Frequency (MHz) Band 1, median noise Level (Resolution of 122.0 kHz) converted to PSD in dBm/Hz P E N CM Figure 65: Median PSD Noise found by Frequency over Band 1 for Different Rx Ports Figure 66 to figure 69 show the minimum, 20 %, median, 80 % and maximum noise PSD presented on the composite band by frequency for each Rx port. ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 58 0 10 20 30 40 50 60 70 80 90 100 -170 -160 -150 -140 -130 -120 -110 -100 -90 -80 -70 Frequency (MHz) Composite band, Rx port: P Level (Resolution of 122.0 kHz) converted to PSD in dBm/Hz max Noise 80% Noise med Noise 20% Noise min Noise Figure 66: Min., 20 %, Median, 80 % and Max PSD Noise for Rx Port P over the Composite Band 0 10 20 30 40 50 60 70 80 90 100 -170 -160 -150 -140 -130 -120 -110 -100 -90 -80 -70 Frequency (MHz) Composite band, Rx port: E Level (Resolution of 122.0 kHz) converted to PSD in dBm/Hz max Noise 80% Noise med Noise 20% Noise min Noise Figure 67: Min., 20 %, Median, 80 % and Max Noise PSD for Rx Port E over the Composite Band ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 59 0 10 20 30 40 50 60 70 80 90 100 -170 -160 -150 -140 -130 -120 -110 -100 -90 -80 -70 Frequency (MHz) Composite band, Rx port: N Level (Resolution of 122.0 kHz) converted to PSD in dBm/Hz max Noise 80% Noise med Noise 20% Noise min Noise Figure 68: Min., 20 %, Median, 80 % and Max Noise PSD for Rx Port N over the Composite Band 0 10 20 30 40 50 60 70 80 90 100 -170 -160 -150 -140 -130 -120 -110 -100 -90 -80 -70 Frequency (MHz) Composite band, Rx port: CM Level (Resolution of 122.0 kHz) converted to PSD in dBm/Hz max Noise 80% Noise med Noise 20% Noise min Noise Figure 69: Min., 20 %, Median, 80 % and Max Noise PSD for Rx Port CM over the Composite Band ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 60 Table 27 compares where the noise PSD was measured, the size of the flat or building and the number of sweeps recorded for each location. In addition, the minimum, median and maximum noise PSD recorded on Band 1 is provided. If there is e.g. a high maximum noise value an excellent radio broadcast reception can be expected in this location. Table 27: Min., Median and Max. PSD Noise Value, Size and No. of Sweeps at each Location, Independent of Rx Port, for Band 1 Location Country Min noise PSD in dBm/Hz 20 % noise Median noise PSD in dBm/Hz 80 % noise Max noise PSD in dBm/Hz Size of flat or building in m² No of sweeps Duerrbachstr Germany -156,48 -154,04 -146,56 -139,18 -95,63 115 64 ImGeiger Germany -155,76 -152,85 -146,85 -135,6 -93,91 130 16 Nauheimerstr Germany -156,6 -155,94 -152,96 -143,15 -100,47 95 16 Rothaldenweg Germany -150,18 -146,62 -143,54 -135,22 -79,83 122 32 VickiBaumWeg Germany -150,44 -144,83 -139,78 -132,15 -91,61 150 48 Calicanto Spain -158,09 -151,39 -146,77 -137,91 -93,47 200 48 Paiporta Spain -157,84 -150,53 -144,58 -132,11 -78,39 96 48 Valencia_46012 Spain -146,98 -137,78 -135,83 -124,45 -81,19 78 48 Valencia_46015 Spain -158,01 -153,8 -148,82 -135,94 -104,25 87 47 Xirivella Spain -158,05 -151,76 -143,79 -132,39 -81,43 110 48 Bourg-La-Reine France -144,79 -143,42 -140,69 -126,99 -92,37 85 112 Spidcom-Lab France -148,22 -143,58 -138,9 -130,72 -92,37 180 64 DeGrandRyStrasse Belgium -153,23 -150,9 -144,8 -135,37 -94,56 130 64 Heidhoehe Belgium -149,73 -146,23 -140,23 -139,23 -77,27 160 64 Huette Belgium -154,04 -152,07 -140,27 -132,31 -98,43 110 64 InDenSiepen Belgium -151,7 -140,87 -139,49 -132,67 -80,16 110 64 Simarstrasse Belgium -156,26 -152,3 -148,29 -135,4 -85,87 150 64 AlsdorferStrasse Germany -145,96 -143,94 -139,13 -132,75 -89,51 110 45 Eichelhaeherweg Germany -149,5 -146,92 -139,06 -128,83 -95,07 85 64 Schlossstrasse Germany -161,83 -154,36 -141,1 -135,12 -97,1 120 64 SchurzelterWinkel Germany -153,44 -150,6 -143,06 -136,06 -98,54 150 64 Wasserkall Germany -145,96 -143,18 -138,26 -131,19 -85,88 64 ColchesterDrive United Kingdom -156,15 -148,87 -145,19 -134,38 -89,7 145 64 WilliamsRoad United Kingdom -157,38 -151,94 -146,57 -134,56 -102,8 97 64 WindmillAvenue United Kingdom -150 -144,73 -141,66 -134,84 -93,2 170 64 WindsorClose United Kingdom -164,23 -159,93 -156,76 -146,01 -106,8 75 64 Cavan France -162,5 -144,43 -141,11 -132,48 -103,17 66 80 Guingamp France -155,94 -144,31 -140,11 -134,66 -79,11 135 108 RueBunuel France -157,1 -153,22 -145,68 -136,62 -96,03 120 80 RueDepasse France -156,15 -146,36 -142,73 -133,45 -94,12 120 96 Trebeurdun France -156,73 -148,36 -142,09 -133,11 -93,75 56 96 ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 61 50 100 150 200 -170 -160 -150 -140 -130 -120 -110 -100 -90 -80 -70 Location size (m2) Level (Resolution of 122.0 kHz) converted to PSD in dBm/Hz Band 1 Max 80% Median 20% Min Figure 70: Min., median and max. PSD Noise versus Size of Location Figure 70 presents a graphical view of the minimum, medium and maximum noise PSD recorded for each location vs. the size of the location. This graph shows that there is no relationship between the noise level received and the size of location. It seems that the noise level recorded at a given outlet is a function of the close electromagnetic environment of this outlet, and is thus not affected by the size of the electrical network or the number of connected appliances. Figure 71 gives the median noise PSD measured on the composite band vs. frequency for all outlets and all Rx ports, separated by country. The noise PSD measured in France, Germany and UK is similar. The noise PSD recorded in Spain presents the highest level of low frequencies below 10 MHz, but is among the lowest values for the band 40 MHz to 80 MHz. Measurements performed in Belgium are the ones with the lowest level for all frequencies from 2 MHz to 80 MHz. However, the electrical network seems more sensitive to FM signals in Belgium than in other countries. ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 62 0 10 20 30 40 50 60 70 80 90 100 -170 -160 -150 -140 -130 -120 -110 -100 -90 Frequency (MHz) Composite band, median noise Level (Resolution of 122.0 kHz) converted to PSD in dBm/Hz Belgium France Spain UK Germany Figure 71: Median Noise PSD found over Frequency over the Composite Band for Different countries |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 7.3.2 Time Domain Noise Statistics | Noise measurements were performed in time domain using a DSO. It is possible to analyze detailed time domain characteristics of noise received. Among the collection of measurements, several typical noise shapes were observed as presented in the following images. For a number of measurements, such as the location presented in figure 72, no variations of the noise characteristics could be observed over the 20 ms observation window. One can identify this typical observation as stationary noise, with no periodicity with respect to the 50 Hz electrical signal. Other measurements present a periodical structure, such as the one illustrated in figure 73. In this typical example, a given noise pattern is reproduced every 10 ms, with a 100 Hz repetition rate. This noise sample is synchronous with the 50 Hz mains period. It can be explained by the periodical variation in impedance at different branches of the network, hence causing a structured noise pattern. In other examples, the noise structure is not synchronous with the 50 Hz mains period. For instance, figure 74 presents a noise sample with an impulsive structure, but the noise pattern is not reproduced regularly. This typical structure can be called asynchronous impulsive noise. Finally, figure 75 presents an example of strong impulsive noise, where very short impulses are repeated periodically. The duration between successive impulses is about 1 ms, leading to a repetition rate of 1 kHz. Such noise can be caused by electronic appliances or energy saving lamps for instance. A quick subjective assessment of the relatively small number of recorded data of all noise samples measured on band 1 provides the following noise occurrence probability: • 60 % of the noise records present no periodicity (e.g. figure 72) • 15 % of the noise records present a periodical, synchronous structure with 50 Hz period (e.g. figure 73) • 15 % of the noise records present an impulsive, asynchronous structure with 50 Hz period (e.g. figure 74) ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 63 • 10 % of the noise records present a strong impulsive structure (e.g. figure 75) 0 2 4 6 8 10 12 14 16 18 20 -10 -5 0 5 10 15 Time (ms) Voltage (mV) Valencia_46015, Band 1 P E N CM Figure 72: Time Domain Noise Record of a Location, No Periodicity -10 -8 -6 -4 -2 0 2 4 6 8 10 -15 -10 -5 0 5 10 15 Time (ms) Voltage (mV) Duerrbachstrasse, Band 1 P E N CM Figure 73: Time Domain Noise Record of a Location, Periodical Structure Synchronous with 50 Hz Period ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 64 0 2 4 6 8 10 12 14 16 18 20 -15 -10 -5 0 5 10 15 Time (ms) Voltage (mV) Wasserkall, Band 1 P E N CM Figure 74: Time Domain Noise Record of a Location, Impulsive Structure, Asynchronous with 50 Hz 0 2 4 6 8 10 12 14 16 18 20 -40 -30 -20 -10 0 10 20 30 40 Time (ms) Voltage (mV) Rue Bunuel, Band 1 P E N CM Figure 75: Time Domain Noise Record of a Location, Strong Impulsive Noise with 1 kHz Period ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 65 Interestingly, the periodic or impulsive nature of the noise samples can be linked to the frequency, due to the availability of successive measurements performed in different bands. Figure 76 compares the time domain noise records for a given location in two frequency bands: Band 1 (2 MHz to 100 MHz) and Band 4 (30 MHz to 80 MHz). In Band 1, one can observe the periodic appearance of noise at a higher absolute level. In Band 4, not only the measured noise level is much lower (note the change in y-axis scale), but the periodic shape of the noise completely disappears. Hence, the bursts of strong noise level observed in Band 1 probably correspond to electromagnetic ingress received in the SW or FM broadcasting bands. 0 5 10 15 20 -20 -15 -10 -5 0 5 10 15 20 Time (ms) Voltage (mV) Xirivella, Band 1 0 5 10 15 20 -4 -3 -2 -1 0 1 2 3 4 Time (ms) Xirivella, Band 4 P E N CM Figure 76: Comparison of Time Domain Noise Records of a Location, for Band 1 and Band 4 In the following, we will consider the correlation between noise samples received on a given band at two different Rx ports. The correlation is characterized by the correlation parameter ρi,j, computed from two time domain noise samples ni(t) and nj(t) as: ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ⎟⎠ ⎞ ⎜⎝ ⎛ − − − = 2 2 2 2 , t n t n t n t n t n t n t n t n j j i i j i j i j iρ (Eq. 1) where denotes the time domain average. Figure 77 and figure 78 give the cumulative probability of the absolute value of ρ for all Rx ports combinations, for Band 1 and Band 4 respectively. As a first observation, the correlation coefficient spans the entire range from 0 (fully decorrelated) to 1 (fully correlated), for all Rx ports combinations. From Band 1, one can observe that Rx ports P and E present the highest degree of correlation, while Rx ports E and CM present the lowest degree of correlation. This difference is less obvious for Band 4, where the CDF move closer together. In general, the correlation coefficient is smaller by roughly 0,1 in Band 4 when compared to Band 1. From this, one can conjecture that signals received via SW and FM bands tend to increase the correlation of noise recorded over different Rx ports. ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 66 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 uncorrelated ← Correlation coefficient |ρ| → correlated Cumulative Probability of correlation coefficient Band 1 P vs E P vs N P vs CM E vs CM N vs CM E vs N Figure 77: Cumulative Probability of Correlation Coefficient for Band 1 Noise Measurements 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 uncorrelated ← Correlation coefficient |ρ| → correlated Cumulative Probability of correlation coefficient Band 4 P vs E P vs N P vs CM E vs CM N vs CM E vs N Figure 78: Cumulative Probability of Correlation Coefficient for Band 4 Noise Measurements ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 67 |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 7.4 Channel Capacity, Spatial Correlation and Singular Values | |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 7.4.1 Singular Values and Spatial Correlation | The channel transfer function measurements (S21 measurements, see clause 7.1) are the basis of the singular value and spatial correlation considerations. The transmit and receive ports used define the MIMO-PLC channel (see left side of figure 79 for two transmit and four receive ports). For each measurement frequency, the complex coefficient hmn from transmit port n (n=1,…,NT) to receive port m (m=1,…,NR) define the channel matrix H of this frequency: ⎟ ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎜ ⎝ ⎛ = T R R T N N N N h h h h L M O M L 1 1 11 H The channel matrix H can be decomposed by means of the singular value decomposition (SVD) H UDV H = , where U and V are unitary matrices (H is the hermitian operator) and D is a diagonal matrix which contains the singular values j λ (j=1,…,r). The number of non-zero singular values depends on the number of transmit and receive ports: r = min(NT, NR). Since D is a diagonal matrix, the MIMO channel is decomposed in r independent SISO paths and the singular values describe the logical paths of the MIMO channel (see right side of figure 79). Figure 79: Channel Matrix and Singular Value Decomposition Figure 80 shows the cumulative probability of the singular values for all measurement sweeps (338 in total) for different MIMO configurations. If only one transmit port is used, the number of logical paths is one. If two transmit ports are used and more than two receive ports are used, two logical paths are available. The solid lines in figure 80 represent the logical path 1 while the dashed lines represent the second logical path. Figure 80 shows the SISO configuration with PN feeding and receiving on P as the SISO reference. However, the SISO configuration with PN feeding and receiving on N shows exactly the same behavior with respect to the cumulative probability. The 2x4 MIMO configuration shows the gain of MIMO compared to SISO (1x1 configuration): At the 50 % point (median) an improvement of 11 dB is provided on the first logical path compared to SISO. The second path adds communication capabilities which are 3 dB less than SISO at the 50 % point. The MIMO gain is even more visible for the highly attenuated channels which are important for meeting coverage requirements. Figure 81 shows a zoomed plot of figure 80 at the 10 % point. The first path of the 2x4 configuration gains 13 dB compared to SISO while the second path provides a decreased communication capacity of 1,5 dB compared to SISO. ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 68 -140 -120 -100 -80 -60 -40 -20 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 singular value [dB] cumulative probability 1x1 (tx: PN; rx: P) 1x2 (tx: PN; rx: P&CM) 1x3 (tx: PN; rx: P&N&E) 1x4 (tx: PN; rx: P&N&E&CM) 2x2 (tx: EP&PN; rx: P&N) 2x3 (tx: EP&PN; rx: P&N&E) 2x4 (tx: EP&PN; rx: P&N&E&CM) stream 1 stream 2 Figure 80: Singular Values -110 -100 -90 -80 -70 -60 0 0.05 0.1 0.15 singular value [dB] cumulative probability 1x1 (tx: PN; rx: P) 1x2 (tx: PN; rx: P&CM) 1x3 (tx: PN; rx: P&N&E) 1x4 (tx: PN; rx: P&N&E&CM) 2x2 (tx: EP&PN; rx: P&N) 2x3 (tx: EP&PN; rx: P&N&E) 2x4 (tx: EP&PN; rx: P&N&E&CM) Figure 81: Singular Values, Zoom of Important Area to show Maximum Coverage ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 69 Figure 82 shows the cumulative probability of the ratio of the singular values. Note that without the loss of generality the first singular value is assumed to be larger than the second singular value. The figure shows the ratio in dB according to: ( ) 2 1 10 / log * 20 λ λ The ratio is calculated for each measurement and frequency point before calculating the cumulative probability. The higher the ratio, the lower the second stream is, indicating a high spatial correlation. The highest spatial correlation is observed for the 2x2 MIMO configuration. The slow slope at high values of the cumulative probability shows that only a few measured channels have a very high spatial correlation. 0 5 10 15 20 25 30 35 40 45 50 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 singular value stream 1 / singular value stream 2 [dB] cumulative probability 2x2 (tx: EP&PN; rx: P&N) 2x3 (tx: EP&PN; rx: P&N&E) 2x4 (tx: EP&PN; rx: P&N&E&CM) Figure 82: Spatial Correlation Figure 82 shows that e.g. for 60 % of the channels, the smallest singular value is 10 times smaller or more than the largest singular value. In radio applications, this would not achieve a doubling of the channel capacity because the SNR is too small. In PLC, due to the large quantity of available SNR, even unbalanced singular values produce a large capacity increase. Additionally, beam forming improves the SNR of the first path. Therefore, PLC performance improvement is not only caused by the 2nd path. This is another difference to WiFi MIMO. |
d41f8dab39a9a600e474dc59c8588181 | 101 562-3 | 7.4.2 Channel Capacity | The decomposition into independent streams described by the SVD is the basis of the channel capacity calculation. The MIMO channel capacity is the sum of the capacity of each single stream [i.7]: s bit N N B C N i j j i T / 1 log 1 1 2 1 , 2 ∑∑ = = ⎟⎟ ⎠ ⎞ ⎜⎜ ⎝ ⎛+ = λ ρ , ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 70 j i, λ is the eigenvalue (squared singular value) of stream j and frequency point i. N is the total number of frequency points and B is the bandwidth (here: 100 MHz). ρ is the ratio of the transmit power spectral density to the noise power spectral density and is assumed to be constant over the frequency, here. ρ = 95 dBm/Hz is used which corresponds e.g. to a transmit power spectral density of -55 dBm/Hz and a noise power spectral density of -150 dBm/Hz at each receive port (this is a quiet environment, roughly 80 % of the noise value of the noise measurements, see figure 61 or figure 63). The noise is assumed to be uncorrelated for each receive port, in this case. The factor 1/NT divides the total transmit power among the available transmit ports, i.e. it is assumed that the total transmit power is the same for SISO and MIMO. The channel capacity gives the theoretical achievable bitrate. Actual modem implementations will not achieve these bitrates, meaning that the limited dynamic range of ADC and DAC does not allow such signal over noise ratios. Figure 83 shows the complementary cumulative probability (C-CDF) of the channel capacity for different MIMO configurations (identical MIMO configurations are used in clause 7.4.1 for the singular values). 0 500 1000 1500 2000 2500 3000 3500 4000 4500 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 channel capacity C [Mbit/s] C-CDF(C), Prob [channel capacity > C] PSD: -55dBm/Hz; NPSD: -150dBm/Hz; 100MHz 1x1 (tx: PN; rx: P); median: 774 Mbit/s 1x1 (tx: PN; rx: N); median: 756 Mbit/s 1x2 (tx: PN; rx: P&CM); median: 856 Mbit/s 1x3 (tx: PN; rx: P&N&E); median: 931 Mbit/s 1x4 (tx: PN; rx: P&N&E&CM); median: 969 Mbit/s 2x2 (tx: EP&PN; rx: P&N); median: 1341 Mbit/s 2x3 (tx: EP&PN; rx: P&N&E); median: 1465 Mbit/s 2x4 (tx: EP&PN; rx: P&N&E&CM); median: 1597 Mbit/s Figure 83: Channel Capacity for Different MIMO Configurations Figure 84 is a zoomed in version of figure 83 at the high coverage point. Here, the gain is better visible compared to the median values. Considering e.g. the 98 % point, SISO achieves 286 Mbit/s while 2x2 MIMO obtains 574 Mbit/s (gain compared to SISO of factor 2) and 2x4 MIMO obtains 759 Mbit/s (gain compared to SISO of factor 2,7). ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 71 200 300 400 500 600 700 800 900 1000 0.86 0.88 0.9 0.92 0.94 0.96 0.98 1 channel capacity C [Mbit/s] C-CDF(C), Prob [channel capacity > C] PSD: -55dBm/Hz; NPSD: -150dBm/Hz; 100MHz Figure 84: Channel Capacity for Different MIMO Configurations: Zoom at High Coverage Point, for key see figure 83 Table 28 summarizes the median values of the channel capacity for the different configurations for each country individually. The table also shows the gain compared to SISO (PN feeding and P receiving). Table 28: Median MIMO Channel Capacity for Different Configurations All Germany Spain France Belgium UK Number of channels 338 119 30 81 60 48 Confi- guration Tx ports Rx ports Mbit/s gain Mbit/s gain Mbit/s gain Mbit/s gain Mbit/s gain Mbit/s gain 1x1 PN P 774 561 787 1 082 692 1 039 1x1 PN N 756 536 766 1 117 676 1 061 1x2 PN P& CM 856 1,11 658 1,17 897 1,14 1 177 1,09 817 1,18 1 095 1,05 1x3 PN P,N,E 931 1,20 687 1,22 952 1,21 1 280 1,18 828 1,20 1 230 1,18 1x4 PN P,N,E, CM 969 1,25 728 1,30 1 006 1,28 1 302 1,20 888 1,28 1 254 1,21 2x2 EP, PN P,N 1 341 1,73 1 019 1,82 1 434 1,82 1 892 1,75 1 177 1,70 1 824 1,76 2x3 EP, PN P,N,E 1 465 1,89 1 123 2,00 1 582 2,01 2 043 1,89 1 274 1,84 2 018 1,94 2x4 EP, PN P,N,E, CM 1 597 2,06 1 230 2,19 1 679 2,13 2 192 2,03 1 440 2,08 2 130 2,05 ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 72 Table 29: MIMO Channel Capacity for Different Configurations at 98 % Coverage All Germany Spain France Belgium UK Number of channels 338 119 30 81 60 48 Confi- guration Tx ports Rx ports Mbit/s gain Mbit/s gain Mbit/s gain Mbit/s gain Mbit/s gain Mbit/s gain 1x1 PN P 287 264 498 457 381 350 1x1 PN N 284 246 513 466 377 407 1x2 PN P& CM 402 1,40 372 1,41 625 1,26 539 1,18 459 1,21 484 1,38 1x3 PN P,N,E 369 1,29 332 1,26 676 1,36 578 1,26 455 1,20 501 1,43 1x4 PN P,N,E, CM 432 1,50 410 1,55 735 1,48 626 1,37 498 1,31 542 1,55 2x2 EP, PN P,N 565 1,96 453 1,71 960 1,93 829 1,81 751 1,97 823 2,35 2x3 EP, PN P,N,E 642 2,23 546 2,07 1 081 2,17 921 2,01 754 1,98 916 2,62 2x4 EP, PN P,N,E, CM 755 2,62 648 2,45 1 214 2,44 1 034 2,26 849 2,23 1 007 2,88 ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 73 Annex A: Useful Information to perform Field Tests A.1 Equipment List used for Field Tests • All measurements - Cable drum to take power supply from neighbor - Isolation transformer - LISN, AMN and additional mains power filter - Stickers to label the outlets - Ground plane 1 m² • 2 MIMO PLC STF410 universal couplers • S21 channel measurements + S11 reflection measurements - NWA - Calibration kit, BNC Short, Open, 50 Ohm termination and through - Long coaxial cables. Double shielded, long enough to reach all electrical outlets in the house under examination. It is recommended to use RG214 cables due to their low attenuations - CM choke absorber • Spectrum Analyzer to verify any noise ingress signals overloading NWA's input • Noise measurements - DSO - 4* HPF-002 High Pass Filter with fc = 1,8 MHz - 4* HPF-025 Mini-circuit SHP-25+ High Pass Filter with fc = 25 MHz - 4* LPF-100 Mini-circuits SLP-100+ Low Pass Filter with fc = 100 MHz - 4* SBF-FM 4 Mini-circuits NSBP-108+ Stop Band Filter rejecting the FM band - AMP BOARD Mini-circuits ZFL-500LNB+ (4x) Board including 4 low noise amplifiers - PWR SUPPLY Team dependent Supplying DC 15 V and 240 mA - DSO Team dependent Digital Sampling Oscilloscope - CABLE up to 12 Cable with 2 male SMA connectors - Adapters from N-connector, BNC, SMA to all combinations ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 74 A.2 Check Lists to Monitor progress at field tests Following Check lists may be filled during field tests. This helps to memory not to miss any record. Measurement Protocol for STF410 MIMO PLC measurements Network Analyser settings: Frequency range: 1MHz -> 100MHz, Maximum number of measuremtn points: 1601, max. feeding Power: 10dBm, min acce save Frequ, Real, Imag in CSV file. Location (Country, Town, zip-code, street, number, location and size in m² of the flat/house under test): Take a photograph of each connected outlet and from the building. Channel Measurements: Through Calibratation (S21) of NWA at the endings of the long cables. Both cables shortcutted Channel Measurements Naming convention: Pt_F_Pr_R_xx.xx.CSV for feeding at plug No t to port F={'EP', 'PN', 'NE'} and receiving at plug No r from por ‘xx.xx’ is the timing distance to the rising LCZC at Tx probe in ms when the sweep was recorded. If trigger of NWA was not in sync to LCZC ‘xx.x Receive P1 P2 P3 P P P P P P Phase (L1, L2 or L3): P E N CM P E N CM P E N CM P E N CM P E N CM P E N CM P E N CM P E N CM P E N CM Feed Feed EP EP P1 PN P1 PN NE NE APN APN PNE PNE EP EP P2 PN P2 PN NE NE APN APN PNE PNE EP EP P PN P3 PN NE NE APN APN PNE PNE EP EP P PN P PN NE NE APN APN PNE PNE EP EP P PN P PN NE NE APN APN PNE PNE EP EP P PN P PN NE NE APN APN PNE PNE EP EP A.3 Data Format to save Recorded Measurements For better exchange a simple ASCII-Format is used for data storage. For data measured with the network analyzer, some header lines indicating the column settings, followed by a table with the measurement results. Its form looks like: # Channel 1 # Trace 1 Frequency, Formatted Data, Formatted Data 1000000, 2.109846e-002, 2.714015e-002 ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 75 1061875, 2.740054e-002, 2.134098e-002 1123750, 2.862551e-002, 9.963092e-003 ... ... ... 99876250, -3.873063e-004, -2.140583e-004 99938125, -3.182144e-004, -2.049750e-004 100000000, -3.223578e-004, -1.644584e-004 The values in the table are: frequency in Hz, S21 real part, S21 imaginary part ETSI ETSI TR 101 562-3 V1.1.1 (2012-02) 76 History Document history V1.1.1 February 2012 Publication |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 1 Scope | MIMO PLT EMI is a review and statistical analysis which takes into account such matters as earthing variation, country variation, operator differences, phasing and distribution topologies, domestic, industrial and housing types along with local network loading. |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 2 References | References are either specific (identified by date of publication and/or edition number or version number) or non-specific. For specific references, only the cited version applies. For non-specific references, the latest version of the reference document (including any amendments) applies. Referenced documents which are not found to be publicly available in the expected location might be found at http://docbox.etsi.org/Reference. NOTE: While any hyperlinks included in this clause were valid at the time of publication, ETSI cannot guarantee their long term validity. |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 2.1 Normative references | The following referenced documents are necessary for the application of the present document. Not applicable. |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 2.2 Informative references | The following referenced documents are not necessary for the application of the present document but they assist the user with regard to a particular subject area. [i.1] Sartenaer, T. & Delogne, P.: "Powerline Cables Modelling for Broadband Communications", ISPLC 2001, pp. 331-337. [i.2] R. Hashmat, P. Pagani, A; Zeddam, T. Chonavel: "MIMO Communications for Inhome PLC Networks: Measurements and Results up to 100 MHz", IEEE International Symposium on Power Line Communications and its Applications (ISPLC), Rio, Brasil, March 2010. [i.3] A. Schwager: "Powerline Communications: Significant Technologies to become Ready for Integration", Doctoral Thesis at University of Duisburg-Essen, May 2010. [i.4] ETSI TR 102 175 (V1.1.1): "PowerLine Telecommunications (PLT); Channel characterization and measurement methods". [i.5] ETSI TR 101 562-1 (V1.3.1): "Powerline Telecommunications (PLT); MIMO PLT; Part 1: Measurement Methods of MIMO PLT". [i.6] ETSI TR 102 616 (V1.1.1): "PowerLine Telecommunications (PLT); Report from PlugtestsTM 2007 on coexistence between PLT and short wave radio broadcast; Test cases and results". [i.7] ITU-R Recommendation BS.1284: "General methods for the subjective assessment of sound quality". [i.8] SCHWARZBECK MESS - ELEKTRONIK; EFS 9218: "Active Electric Field Probe with Biconical Elements and built-in Amplifier 9 kHz ... 300 MHz". NOTE: See http://www.schwarzbeck.de/Datenblatt/m9218.pdf. [i.9] ETSI TR 101 562-3 (V1.1.1): "PowerLine Telecommunications (PLT); MIMO PLT; Part 3: Setup and Statistical Results of MIMO PLT Channel and Noise Measurements". ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 7 [i.10] R&S®HFH2-Z2: "Loop Antenna Broadband active loop antenna for measuring the magnetic field-strength; 9 kHz - 30 MHz". NOTE: See http://www2.rohde- schwarz.com/en/products/test_and_measurement/emc_field_strength/emc_accessories/. [i.11] CISPR 11 (Ed. 5.0): "Industrial, scientific and medical equipment - Radio-frequency disturbance characteristics - Limits and methods of measurement". [i.12] CISPR 22 (Ed. 6.0): "Information technology equipment - Radio disturbance characteristics - Limits and methods of measurement". |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 3 Symbols and abbreviations | |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 3.1 Symbols | For the purposes of the present document, the following symbols apply: A or Att Attenuation in dB E Electrical Field strength in dBµV/m H Magnetic field in dBµA/m k Coupling factor in dB(µV/m)-dBm P Power in dBm PSD Power Spectral Density in dBm/Hz sxy Scattering parameter in dB U Voltage in dBµV |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 3.2 Abbreviations | For the purposes of the present document, the following abbreviations apply: AF Antenna Factor AM Amplitude Modulation ASCII American Standard Code for Information Interchange BNC Bayonet Nut Connector CDF Cumulative Distribution Function CM Common Mode CSV Comma Separated Values DC Direct Current DM Differential Mode E Protective Earth Contact EMC Electromagnetic Compatibility EMI Electro Magnetic Interference FD Frequency Domain FM Frequency Modulation FTP File Transfer Protocol GPS Global Positioning System HF High Frequency HIFI High Fidelity IF Intermediate Frequency LCZC Line Cycle Zero Crossing LISN Line Impedance Stabilization Network LVDN Low Voltage Distribution Network MIMO Multiple Input Multiple Output N Neutral Contact NOTE: Used as decoupling filter. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 8 NWA Network Analyser P Phase or Live Contact PC Personal Computer PE Protective Earth PLC PowerLine Communication PLT PowerLine Telecommunications PSD Power Spectral Density RF Radio Frequency Rx Receiver SINPO Signal, Interference, Noise, Propagation, Overall SISO Single Input Single Output STF Special Task Force TD Time Domain Tx Transmitter VHF Very High Frequency |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 3.2.1 Abbreviations used for feeding styles | APN Signal feed mode: Dual wire feed (version C of clause 7.1.4.5 in [i.5]) to input P||N E in figure 28 in [i.5] CM Signal feed mode: Common mode, P, N, E terminated to ground (CM is expected to be used only for receiving by PLT modems) EP Signal feed mode: DELTA (differential) between E and P, PN and NE terminated EP-NET Signal feed mode: Differential between E and P, only NE terminated EPNT Signal feed mode: DELTA (differential) between E and P, PN and NE not terminated NE Signal feed mode: DELTA (differential) between N and E, PN and EP terminated NE-EPT Signal feed mode: Differential between N and E, only EP terminated NENT Signal feed mode: DELTA (differential) between N and E, PN and EP not terminated PN Signal feed mode: DELTA (differential) between P and N, NE and EP terminated PNE Signal feed mode: Dual wire feed (version C of clause 7.1.4.5 in [i.5]) to input P||N E in figure 28 in [i.5] PNNT Signal feed mode: DELTA (differential) between P and N, NE and EP not terminated (SISO) |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 4 Major Project Phases | Table 1 No. Period Topic Event 01 Sept. 2010 Project organization Definition of targets, what and how to measure STF 410 Preparatory Meeting Stuttgart, Germany 02 Nov 2010 Setup of MIMO PLT measurements (EMI, Channel and Noise) Several STF 410 phone conferences. Drafting of measurement specification 03 Dec. 2010 1st version of the STF410 couplers Coupler to send and receive MIMO PLT signals developed 04 Jan 2011 and later Verification of couplers and filters developed for STF410.14 identical couplers are manufactured and shipped to the STF experts Couplers are used by STF410 experts in field measurements in private homes 05 March 2011 Agreement on STF410 logistics, when and where to perform field measurements 06 April 2011 Approval of 1st TR on STF410 couplers ETSI PLT#59 07 March 2011 to June 2011 Field measurements in Spain, Germany, France, Belgium and the United Kingdom 08 June 2011 Statistical evaluation of results Several STF 410 phone conferences 09 July 2011 Approval of 2nd TR on EMI results ETSI PLT #60 10 Oct. 2010 to August 2011 Evaluation of worldwide presence of PE wire 11 June 2011 to August 2011 Drafting and STF 410 review and approval process ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 9 No. Period Topic Event 12 Sept. 2011 Presentation of channel and noise measurement to ETSI PLT plenary ETSI PLT #61 13 Oct 2011 Revision and rearrangement of TR content for all 3 parts 14 Nov 2012 Approval of all 3 parts of TR 101 562 ETSI PLT #62 |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 5 Motivation | PLT systems available today use only one transmission path between two outlets. It is the differential mode channel between the phase (or live) and neutral contact of the mains. These systems are called SISO (Single Input Single Output) modems. In contrast, MIMO PLT systems make use of the third wire, PE (Protective Earth), which provides several transmission combinations for feeding and receiving signals into and from the LVDN. Various research publications [i.1], [i.2] or [i.3] describe that up to 8 transmission paths might be used simultaneously. Further description of: • motivation for MIMO PLT; • installation types and the existence of the PE wire in private homes; • measurement Setup description to record throughput communication parameters and their results; can be found in [i.5] and [i.9]. |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 6 Measurement Description | |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 6.1 Introduction | EMI properties of the LVDN can be recorded in Time- (TD) or in Frequency Domain (FD). The pros and cons of each measure were evaluated early on by STF 410. It was concluded that the FD approach is better suited for the following reasons. Most of the earlier EMC measurements relating to PLC were performed in FD. Thus the comparison between the results obtained by STF 410 and those of the past is much easier in FD. The human ear is essentially an FD analyser. Interferences assessed by human ears like the SINPO measurements use Consumer Electronic devices like AM or FM radio receivers. Such measurements were performed in [i.6] and [i.7]. MIMO test signals are fed to all Tx paths simultaneously or sequentially. These investigations are conducted with a pulsed signal to allow recognition by the human ear-brain-chain. NOTE: See http://stason.org/TULARC/radio/shortwave/08-What-is-SINPO-SIO-Shortwave-radio.html. Field levels are monitored with a calibrated antenna, which is straight forward to process in FD. EMI measurements in TD have the risk that periodicities in the transmitted PN-sequence may cause additional spurs. Furthermore, the measurement dynamic does not seem to be adequate in TD. EMI principally occurs during transmissions of PLC modems and is considered in statistical evaluations. FD measurements can be done using a comb generator and spectrum (or EMI) analyser. This setup has the benefit that transmitter and receiver do not need to be synchronized. On the other hand the dynamic range or frequency resolution is limited due to the feeding energy of the comb generator needing to be shared among all signal carriers. Alternatively, a sweeping source like a network analyser (NWA) might be used. Special care has to be taken with signals received by the antenna, as they can be influenced by additional signals being picked up through the long cables connecting the antenna to the NWA. To minimize this effect, double shielded cables, common mode absorption devices (CMADs) and ferrites have to be installed. This measurement method has been selected by STF 410 due to the faster recoding time of a frequency sweep and the high dynamic range. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 10 To increase the number of measurements recorded, STF 410 is split into several teams operating in parallel in various countries. Measurement campaigns where conducted in Germany, Switzerland, Belgium, France and Spain. To guarantee comparability of the individually recorded data each team is equipped with identical probes or PLT couplers. The antenna was shipped to each team in turn. The actual measurements were performed with a general purpose NWA. A commercially available, small biconical antenna (with built-in amplifier) was used because of its frequency range of up to 100 MHz. In one location the loop antenna (limited to frequencies up to 30 MHz) is used for a comparison of this field tests with earlier measurement campaigns. Figure 1 shows the measurement equipment used for EMI measurements. Biconical Active Electric Field Probe [i.8] AM, FM radio receiver: Sony® ICF-SW1000T Biconical Antenna on wooden tripod Loop Antenna (magnetic field) [i.10] ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 11 NWA, Spectrum Analyser, Amp, Isolation Transformer, LISN and power filters NWA, Amp, mounted Antenna, and double shielded cables NOTE: Sony® ICF-SW1000T is an example of a suitable product available commercially. This information is given for the convenience of users of the present document and does not constitute an endorsement by ETSI of this product. Figure 1: Measurement Equipment Used by Individual Teams |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 6.2 General Requirements for the Measurements | The power supply for measurement equipment has to be prepared prior to starting measurements. The supply should be clean and maximally separated from the grid of the residential unit being tested. It is recommended that the power supply be taken from a neighboring flat, a backup power supply or a least a plug far away from the installation to be assessed. If there is a connection to the electricity grid, the power supply has to be filtered. A filtering device for phase, neutral and the protective earth is documented in [i.5]. Additionally, an isolation transformer is used to filter protective earth as most power filters today do not filter the protective earth wire. This is also true for the embedded filters in the measurement equipment used. The test signals for all EMI measurements are fed using the MIMO PLC couplers specified in [i.5]. |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 6.3 Radiation Measurements (k-factor) | |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 6.3.1 Set-Up | The measurement setup basically consists of a NWA connected with coupler A to the mains. The power supply of the NWA is isolated from the LVDN being tested, by a filter providing CM- and DM impedances, seen from the LVDN, of > 1 kΩ. To enhance the dynamic range of the setup, the NWA is connected to an amplifier and the amplified signal is fed into the MIMO Coupler. On the other side, the antenna is connected through a cable with ferrites to a high-pass and the receiving end of the NWA. The HPF-002 described in [i.9], clause 6.6.1 (Noise Measurement Set-up) can be used as a high-pass filter. It attenuates signals below 2 MHz. In a few cases signals below 2 MHz have been identified, reducing the dynamic range of the NWA. This is why they have to be filtered. For years experts claimed that NWA k-factor measurements using coaxial cables to connect the couplers were unacceptable, because of the resulting "loop". Thus the measurement setup described herein was validated by comparative measurements with a setup using a fiber-optical link between the antenna and the NWA. No difference could be detected. Thus, the optical link was not further used, because of its limited dynamic range, higher noise and more cumbersome installation. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 12 Figure 2: General Measurement Set-up for Radiated EMI Figure 3: General Measurement Set-up to Record the k-Factor Outlets used for feeding signals are arbitrarily selected from within the building. The antenna is positioned at a distance of 10 m or 3 m from the exterior wall outside the building. Some antenna points are also selected within the building. Several antenna locations may be selected and the radiation recorded. If the measurement dynamic is not sufficient (signal has to be at least 10 dB above noise floor, i.e. the signal indicated by the NWA without the signal injection connected) an RF amplifier is placed in the line between the NWA generator and the signal injection box. Care should be taken, that the output power does not exceed 1 W to avoid damaging the injection boxes and disturbing the appliances connected to the mains grid. If there is a risk of this happening, an attenuator of 30 dB has to be inserted between the cable connectors for calibration. To calculate the k-factor, the 30 dB has to be subtracted from that derived from Eq.1. NWA is operated using the following settings: • Start Frequency: 1 MHz • Stop Frequency: 100 MHz • Number of measurement points per sweep: 1 601 • IF Bandwidth: 1 kHz • Feeding Power: +10 dBm, 0 dBm • Data are recorded in ASCII format including at least: frequency, Real part, Imaginary part, absolute value in dB. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 13 Care has to be taken that the amplifier is not saturated. The file name convention of the EMI record is: Ptt_Fa_Ayy_Dp_o_xx.xx.CSV where: • 'tt' is the number of the transmitting plug. The 1st digit indicates the level in the building where feeding was done. • 'Fa' is the port where signals are fed differentially: EP, PN, NE, EPNT, PNNT, NENT, APN, PNE, EP-NET, NE-EPT (see figure 6). • 'yy' identifies the location of the antenna (e.g. A01, A02, …., leading zeros are required). • 'p' specified the place of the antenna: '0' is for 10m distance, '3' for 3m distance outside the building and 'I' for indoor. • 'o' is the orientation of the antenna: - 'v' or 'h' in case of the biconical antenna. 'h' means the axis from dipole to dipole is parallel to the horizon and 'v'-direction is vertically. Since this measurement campaign focuses on the radiation produced by PLT, the measurements are performed with these two polarisations in agreement with typical disturbance field strength measurements for products as defined in CISPR 11 [i.11] and CISPR 22 [i.12]. The higher value of the 2 orientations is used as specified in clause 6.3.4. - 'x', 'z' or 'z' in case of the loop antenna (x means H-field parallel to the building wall; z means H-field towards ground). It is common practice to measure the magnetic field in three directions (e.g. see German SchuTSEV). The vector sum of the 3 orientations will give the total H-field as specified in clause 6.3.4. • 'xx.xx' is the timing distance to the rising LCZC at Tx coupler in ms when the sweep was recorded. If trigger of NWA was not in sync with LCZC 'xx.xx' is not applied. E.g. if the filename is P22_PNNT_A01_D3_v.csv the feed was done between P and N in the delta style and the 2 other ports (NE and EP) are not terminated. This is the conventional SISO style. The biconical antenna was located at antenna position 01 in 3 m distance from the outside wall of the building in a vertical orientation. All antenna measurements are saved in the 'EMI' folder of STF410 repository. The folder tree consists of: STF 410 Initials of Expert Name of Location EMI. A ground plane is required, at least for the common mode injection. The ground plane has to be directly connected (low inductance) to the coupling box and be at least 1 m2 in size. For convenience the file handling tool (see annex B) can be used. This tool also can be a helpful guide when reading through the measurements. |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 6.3.2 Calibration of NWA | The NWA needs to be calibrated in order to eliminate the effects caused by the need to use long cables in the building. A response (thru) calibration is done by shortcutting the endings of both coaxial cables. A conventional adapter (BNC female to BNC female) is used as a calibration kit. Prior to starting measurements, the NWA has to be calibrated according to figure 4. To prevent the NWA from being overloaded with input, the NWA generator setting has to be turned down as much as possible (typically -25 dBm). If the output power of the amplifier is still too much for the NWA input, refer to the alternative calibration procedure in annex A. The Analyser will usually automatically correct the calibration data, after the calibration process, when the feeding power is increased. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 14 NWA amplifier cable used for signal injection antenna cable (with ferrites for suppression of sheat current) Figure 4: NWA Calibration During measurements, the cable ends of the NWA have to be connected to the MIMO coupler and the antenna according to figure 5. The generator output power can be increased to improve the dynamic range of the measurements. Care should be taken not to exceed an output power of 1 W, in order to prevent overloading the MIMO coupler. NWA amplifier cable used for signal injection MIMO PLT coupler LV- installation antenna antenna cable (with ferrites for suppression of sheat current) s21 Figure 5: Use of NWA and Set-up for the Measurements ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 15 |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 6.3.3 Signal Injection | For the coupling modes, the following switch settings for the boxes are to be used. Coupling mode Switch setting PNNT DELTA (differential) mode PN, NE and EP NOT terminated (standard SISO PN) (see clause 7.1.4.1 of [i.5]) EPNT DELTA (differential) mode EP, PN and NE NOT terminated (SISO EP) (principle shown in clause 7.1.4.1 of [i.5]) NENT DELTA (differential) mode NE, EP and PN NOT terminated (SISO NE) (principle shown in clause 7.1.4.1 of [i.5]) ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 16 Coupling mode Switch setting PN DELTA (differential) mode PN, NE and EP terminated (MIMO) (principle shown in clause 7.1.4.2 of [i.5]) EP DELTA (differential) mode EP, PN and NE terminated (MIMO) (principle shown in clause 7.1.4.2 of [i.5]) NE DELTA (differential) mode NE, EP and PN terminated (MIMO) (see clause 7.1.4.2 of [i.5]) ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 17 Coupling mode Switch setting EP-NET partial delta type injection, signal between P and E, N-E terminated, P-N not terminated (MIMO) (see clause 7.1.4.3 of [i.5]) (MIMO Asymmetric Transmit) NE-EPT partial delta type injection, signal between N and E, P-E terminated, P-N not terminated (MIMO) (see clause 7.1.4.3 of [i.5]) (MIMO Asymmetric Transmit) APN Dual wire, input P||N - E (see clause 7.1.4.5 (version C) of [i.5]) ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 18 Coupling mode Switch setting PNE Dual wire input PN (see clause 7.1.4.5 (version C) of [i.5]) Figure 6: PLT Coupler Switch Settings The figures shown on the right side of figure 6 are screen shots of the software supporting the measurements. This software is described in annex B of the present document. |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 6.3.4 Calculation of the Final k-Factor | To evaluate the radiation of buildings the coupling factor (k-factor) is defined by: Coupler PLT Coupler PLT output amp ceiver Coupler PLT output amp ceiver feed antenna H E A AF s A P AF P A P AF U P E k _ 21 _ _ max, Re _ _ max, Re max, , dBm) - V (dB 107 dBm) - V (dB 107 + + + = + − + + = + − + = − = μ μ (Eq. 1) with: Eantenna: the field strength received at the location of the antenna, unit: dB(µV/m). Pmax,feed: signal at the output of the PLT coupler (in case of terminated output), unit dBm. Pmax,amp_output: signal at the output of the amplifier provided at the cable end (in case of termination), unit dBm. APLT_Coupler: Attenuation of the PLT coupler as described in [i.5], unit dB. UReceiver: voltage at the output of the antenna, unit dB(µV). PReceiver: power from the output of the antenna, unit dBm. AF: antenna factor of the antenna, unit dB(1/m). s21: scattering parameter as measured by the network analyser with valid calibration, unit dB. NOTE: If the alternative calibration procedure of annex A is used, the corrected s21 values have to be used in Eq. 1. kE,H: k-factor with regard to the electric field component (kE) or magnetic field component (kH), unit dB(µV/m)-dBm. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 19 The k-factor is used first in [i.4]. The formula above says: If a signal is fed with 0 dBm into the mains of a building an electrical field of E dBµV/m is recorded outside the building. From the recorded values s21 of the network analyser, the k-factor can be derived using Eq.1. Depending on the antenna used and the coupling, different values have to be used for APLT_Coupler. Table 2: Coupling Types Coupling type APLT_Coupler EPNT, PNNT, NENT Values taken from clause 7.1.4.1 of [i.5] EP, PN, NE Values taken from clause 7.1.4.2 of [i.5] APN, PNE Values taken from clause 7.1.4.5 of [i.5] EP-NET, NE-EPT Values taken from clause 7.1.4.3 of [i.5] The combinations of different antenna polarisations or orientations are antenna dependent. The following calculations apply to derive a single k-factor per injection-plug - antenna location combination. Table 3: Calculation of Resulting k-Factor in Dependence of Antenna Type Antenna type Calculation of the resulting k-factor biconical ( ) vertical horizontal res k k k , max = loop 2 2 2 z y x res k k k k + + = These calculations are performed individually for each frequency in each record. 6.4 Subjective Evaluation of the Interference to Radio Broadcast |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 6.4.1 General | Subjective evaluations of interference to AM radio reception in the HF bands were performed by ETSI STF 332 (PlugtestsTM on coexistence between PLT and short wave radio broadcast) and are documented in [i.6]. Performing identical tests with all MIMO feeding possibilities would deliver unstable results, because the variance of received signal level (fading in time domain) is more dynamic than an operator might be able to test. During a MIMO test, the interference from all MIMO feeding possibilities should be compared. The signal level is usually never stable in HF bands. [i.3] describes dynamic changes in the HF signal level received caused by reflections on the ionosphere. Broadcasting conditions in VHF are by far more stable over time, allowing a comparison of levels recorded over a period of a few minutes. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 20 LVDN R&S SMY01 FM-Signal Modulator Portable PC Audio Mains Filter Step Attenuator MIMO PLT Coupler Ext. Power Supply CM Filter Figure 7: Basic Set-up for FM Interference Tests How to feed the interference signal to the mains is described in clause 6.3.3. The source of the signal is a broadband noise generator or frequency generator with the option of frequency modulation with variable frequency excursion (e.g. Rohde&Schwarz® SMY, see note 2). This generator is modulated with a noise signal. NOTE 1: The noise signal can be generated via sound output from a laptop or PC using the scope software (http://www.zeitnitz.de/Christian/scope_de). NOTE 2: Rohde&Schwarz® SMY is an example of a suitable product available commercially. This information is given for the convenience of users of the present document and does not constitute an endorsement by ETSI of this product. The 3 dB-bandwidth of the modulated signal is at least 240 kHz for the evaluation of radio interference in the ultra short wave bands (FM-Bands). The generator can be switched on and off, in order to distinguish the disturbed and undisturbed states. A blanking input controlled with a rectangular signal (a few Hertz) is preferable. A sweep in the shape of a sine wave can be inserted into the audio signal, which is to be FM modulated, with the use of a software tool like the one described above. The sweeping tone makes it easier to single out the source of this interference from several interferences when listening to sensitive FM broadcastings, as the tone can be detected by the human ear. The FM modulated signal will be injected as an interfering signal to the mains. An example of the spectrum realized with the noise source for the FM-Bands is shown in figure 8. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 21 0 10 20 30 40 50 60 70 99 99.5 100 100.5 101 U in dB(uV) frequency in MHz Figure 8: Output Spectrum obtained with the SMY Generator (frequency modulated with noise) obtained with a Spectrum Analyzer at Resolution Bandwidth of 200 Hz in Clear-Write Modus |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 6.4.2 Verification and Calibration | Prior to the test, the disturbance signal has to be analyzed with a measurement receiver or spectrum analyzer to document the 3 dB bandwidth. As a first step, the amplification between the generator output level and the signal injected into the BNC-plug of the MIMO PLT coupler has to be determined. This is done by connecting the feed cable via an attenuator of 30 dB (protection of the measurement receiver) to the measurement receiver input. Setting of the measurement receiver: Detector: Average Bandwidth: 120 kHz for measurement frequency above 30 MHz Attenuation: Auto Measurement time: 1 000 ms Feeding level of signal generator is Umax_feed. |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 6.4.3 Measurement Procedure | After the calibration of the amplification has been done, the measurement can be performed. The output of the RF generator is connected via a step attenuator to the MIMO PLT coupler (see figure 7). In preparation a couple of radio stations, which can be received at the receiver's location, are selected. At each frequency of a selected radio station the level of the RF generator is adjusted to a lower level where no disturbance at the radio receiver is recognized. (This could be performed easily by e.g. rotating the knobs of the step attenuator to AttSA.) From that value, the generator level is increased until a disturbance can barely be recognized. After that the level of interference signal is verified by connecting Ugen = Umax_feed - AttSA (Eq. 2) to the measurement receiver input to measure the signal level. The measured value Ugen (in dB(µV)) is recorded in an Excel sheet. This procedure is repeated for each: • coupling type; • selected frequency; ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 22 • feeding outlet (injection point); • radio receiver; • radio receiver location. Furthermore the measurements have to be done when the radio receiver is battery driven and mains powered. Figure 9: Radio Reception in a Building with Noise Feeding |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 6.5 General Equipment List | |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 6.5.1 Coaxial Cables | The coaxial cables used to record k-factor10m measurements are doubly shielded (e.g. RG214 are recommended due to their low attenuation). To avoid signal ingress to the cable going back from the antenna or the PLT coupler to the NWA, the cable has to be surrounded by ferrites. Axial Ferrite Beads are attached to the coaxial cable every 0,15 m. For years experts claimed that NWA k-factor measurements using coaxial cables to connect the couplers were unacceptable, because of the resulting "conducting" loop. Thus we validated the measurement setup described herein by comparative measurements with a setup using a fiber-optical link between the receiving coupler and the NWA. No difference could be detected. Thus, the optical link was not further used, because of its limited dynamic range, higher noise and more cumbersome installation. Figure 10 shows a coaxial cable, Ecoflex 10 (double shielded and connected to the mains coupler), equipped every 0,15 m with attached Suppression Axial Ferrite Beads (Würth-Elektronik part number: 74270056). An RG214 cable (black colour) is also visible. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 23 Figure 10: Cable with Ferrites |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 6.5.2 Network Analyzer | See [i.5] for the list of NWA used for the EMI measurements. |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 6.5.3 Probes to Connect to the LVDN | The MIMO PLC couplers for feeding and receiving signals are specified in [i.9]. |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 6.5.4 Amplifier | The Amplifier used to increase the measurement dynamic is: • 50WD1000 (DC - 1 GHz, AR); • Bonn Elektronik BLWA 0310-1. NOTE: Bonn Elektronik BLWA 0310-1 is an example of a suitable product available commercially. This information is given for the convenience of users of the present document and does not constitute an endorsement by ETSI of this product. Figure 11: Amplifier ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 24 |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 6.5.5 Filter to Isolate Measurement Devices from Mains | Filter as specified in clauses 6.4 and 6.5 (Mains Filter) of [i.5] is used. |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 7 Statistical Evaluation of Results | |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 7.1 k-Factor | STF 410 measured the k-factor at 15 locations in Spain, France and Germany. A typical sweep from 1 MHz to 100 MHz of any k-factor measurement is shown in figure 12. Fading characterizes the shape of a k-factor sweep. In total 1,294 such sweeps were recorded by STF 410. 0 1 2 3 4 5 6 7 8 9 10 x 10 7 0 10 20 30 40 50 60 70 Frequency k-factor in dBµV/m-dBm Figure 12: Typical Sweep of a k-Factor Measurement Outdoors at 10 m Distance Figure 13 shows the median of all data separated into the individual feeding possibilities. The median value for each measured frequency and feeding style is calculated individually. NOTE: This median value is derived from data received from all antenna locations: indoors, outdoors in 3 m and 10 m distance from the building. Figure 13 shows no tendency of the k-factor over frequency for all symmetrical feeding possibilities. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 25 The attenuation, caused by terminations of the coupler feeding ports, is not considered in the present document. In all the following graphs the symmetrical feedings have individual terminations, as described in [i.5]. E.g. when feeding EPNT, NENT and PNNT the unused MIMO feeding interfaces are not terminated. When feeding EP, NE and PN the 2 unused feeding possibilities at the PLT coupler are terminated with 50 Ω. At EP-NET and NE-EPT one of the two unused feeding possibility is terminated and the other one is not. The purpose of this presentation is to allow a comparison of the SISO style (PNNT) which is used by conventional PLT modems. The energy feed into the mains after the PLT coupler is 1,3 dB less in the 3-port termination (PN, EP and NE) compared to 1-port termination (EPNT, NENT and PNNT). 0 1 2 3 4 5 6 7 8 9 10 x 10 7 40 45 50 55 60 65 70 Frequency Median k-factor (dBµV/m - dBm) All feedings vs. frequency _EPNT_ _EP_ _NENT_ _NE_ _PNNT_ _PN_ _APN_ _PNE_ _EP-NET_ _NE-EPT_ Figure 13: Median Values for Each Feeding Possibility Over Frequency The legend for figure 13 shows the feeding possibilities according to figure 6. Table 4 wraps up the abbreviations as a reminder. Table 4: Legend of plots EPNT Signal feed mode: DELTA (differential) between E and P, PN and NE not terminated EP Signal feed mode: DELTA (differential) between E and P, PN and NE terminated NENT Signal feed mode: DELTA (differential) between N and E, PN and EP not terminated NE Signal feed mode: DELTA (differential) between N and E, PN and EP terminated PNNT Signal feed mode: DELTA (differential) between P and N, NE and EP not terminated (SISO) PN Signal feed mode: DELTA (differential) between P and N, NE and EP terminated APN Signal feed mode: Dual wire feed (version C in [i.5]) to input P||N E in figure 28 of [i.5] PNE Signal feed mode: Dual wire feed (version C in [i.5]) to input PN in figure 28 of [i.5] EP-NET Signal feed mode: Differential between E and P, only NE terminated NE-EPT Signal feed mode: Differential between N and E, only EP terminated The median k-factor value of each location is given in table 5 for all Antenna positions indoors, at 3 m distance and at 10 m distance from the outside wall of the building. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 26 Table 5: Median Coupling Factors of Each Location Location Country Median k-factor indoor in dBµV/m - dBm Median k-factor from 3 m distance in dBµV/m - dBm Median k-factor from 10 m distance in dBµV/m - dBm Duerrbachstr Germany 72,60 not measured 45,63 ImGeiger Germany 69,17 not measured 44,77 Nauheimerstr Germany 73,48 not measured 43,24 Rothaldenweg Germany 73,60 not measured 51,54 Schlossbergstr Germany 68,46 57,10 44,74 VickiBaumWeg Germany 61,19 not measured 47,70 Boenen Germany 71,88 62,38 not measured Universitaet Germany not measured 55,58 49,84 Voerde Germany not measured 69,53 59,76 El_Puig Spain 55,89 40,04 not measured Sant_Sperit Spain not measured 49,51 44,87 Torre_en_Conill Spain 57,83 45,12 31,87 Guingamp France 72,92 59,52 54,22 RueBunuel France 69,61 not measured 52,67 RueDepasse France 70,67 62,89 50,69 All locations 67,98 51,3 46,96 Due to the high variance of the results between the individual locations and the low number of locations surveyed, a statistical evaluation of the k-factor for each country has not calculated. Furthermore, the number of records at each location is unique. The number of antenna positions was selected according to the size of the location, size of the garden and accessibility to each location. In total, in all frequencies and feeding possibilities: 771,682 values (482 sweeps) have been recorded from a 10 m distance; 650,006 values (406 sweeps) from 3 m; and 441,876 values (276 sweeps) from indoors. An explanation why there is such a high spread in the median values might be due to the local conditions surrounding the building and inside the flat. In most measurements conducted the area around the building was flat and the outdoor antenna positions were located on the same level as the ground floor. If the residential unit was e.g. a flat located in the 2nd floor of a building or a multi level house some of the feeding outlets have an additional vertical distance to the antenna. For example, the k-factor measurements in France and Voerde were recorded where all feeding outlets were located on the ground floor and the area around the building is flat land. This is why the outdoor k-factors in these locations tend to be higher than at others. The cumulative distribution of the k-factor at a location where all 3 antenna positions were recorded with a high number of sweeps is depicted in figure 14. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 27 0 10 20 30 40 50 60 70 80 90 100 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Cumulative Probability of k-factor k-factor [dBµV/m - dBm], independent of Frequency ( 1 - 100MHz) or Link Torre_en_Conill indoor 3m outdoor 10m outdoor NOTE: No. of sweeps indoor: 30, in 3 m: 60, in 10 m: 30. Figure 14: Cumulative Distribution of k-Factor of a Location For MIMO feedings the potential of interference relative to the SISO case is of interest. Figure 15 and figure 16 show the change in k-factor when using the SISO style and other feeding possibilities. The SISO case is the feeding style used by conventional (non MIMO) PLT modems. Figure 16 is a zoom into figure 15 at the median value (50 % point). The lines presented in figure 15 are calculated by subtracting the k-factor of the PNNT from all other k-factor measurements. For the PNNT feeding this subtraction has to result in zero. This is why the PNNT line is presented as a vertical line here. This calculation was performed only where feeding was done at the identical outlet and the signals were received at identical antenna positions for both records. The signals feed in PN and PNE style show a high correlation to PNNT. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 28 -30 -20 -10 0 10 20 30 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Distance to SISO feed [dB], Outlet or Ant-Pos always match CDF _EPNT_ _EP_ _NENT_ _NE_ _PNNT_ _PN_ _APN_ _PNE_ _EP-NET_ _NE-EPT_ Figure 15: Relative Difference of MIMO Feeding Possibilities to SISO Feeding (PN and Others Non Terminated) ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 29 The zoom in figure 16 and table 6 show that only the EPNT feeding style tends to display a higher radiation than the traditional SISO feeding. All other differential feeding possibilities radiate less. -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Distance to SISO feed [dB], Outlet or Ant-Pos always match CDF _EPNT_ _EP_ _NENT_ _NE_ _PNNT_ _PN_ _APN_ _PNE_ _EP-NET_ _NE-EPT_ Figure 16: Zoom Showing the Difference between SISO and MIMO Feedings Table 6 presents the median values of figure 15 and figure 16 in numbers. Table 6: Median Difference at k-Factor of Feeding Style to SISO Feeding Style Difference to SISO feed in dB EPNT 0,89 EP -0,32 NENT -0,21 NE -1,54 PNNT 0,00 PN -1,11 APN -0,96 PNE -1,09 EP-NET -0,11 NE-EPT -0,10 Comparisons between the magnetic field (H-field) and electric field (E-field) were recorded at the location in Voerde, Germany. Radiation measurements were recorded from the building using an E-filed biconical antenna [i.8] and H-field ring antenna [i.10], from 3 m and 10 m away, with the antennae in the same fixed position for each reading. In order to compare H-field and E-field values, the magnetic fields - recorded in dBµA/m – need to be converted into electric fields with a free space wave impedance of 377 Ω = 51,5 dBΩ. Figure 17 and figure 18 show some correlation between H- and E-field at 3 m distance. Obviously, a distance of 3 m may still be in the near field, where the free space wave impedance of 377 Ω cannot be applied. From 10 m away from the building, the E- and H- fields display a similar pattern, as expected in the far field (see figure 19 and figure 20). The graphs stop at 30 MHz because the loop antenna [i.10] is only specified up to this frequency. At frequencies above 30 MHz it is expected that far field radiation conditions from a building are valid at closer distances or even indoors. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 30 The k-factor records in the present document were obtained using an E-field antenna, because magnetic EMC antennae are not available frequency ranges up to 100 MHz. Furthermore, consumer electronic devices in private homes use an E-filed antenna (stick or whip) in the HF and VHF bands. 0 0.5 1 1.5 2 2.5 3 x 10 7 0 10 20 30 40 50 60 70 80 90 100 Frequency Voerde, 3m Antenna Distance, E-Field k-factor (dBuV/m - dBm) 80% k-factor med k-factor 20% k-factor Figure 17: k-Factor Measured with Biconical Antenna at 3 m Distance ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 31 0 0.5 1 1.5 2 2.5 3 x 10 7 0 10 20 30 40 50 60 70 80 90 100 Frequency Voerde, 3m Antenna Distance, Converted from H-Field using 377Ohm k-factor (dBuV/m - dBm) 80% k-factor med k-factor 20% k-factor Figure 18: k-Factor Measured with Loop Antenna in 3 m Distance and Converted to E-field 0 0.5 1 1.5 2 2.5 3 x 10 7 0 10 20 30 40 50 60 70 80 90 100 Frequency Voerde, 10m Antenna Distance, E-Field k-factor (dBuV/m - dBm) 80% k-factor med k-factor 20% k-factor Figure 19: k-Factor Measured with Biconical Antenna at 10 m Distance ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 32 0 0.5 1 1.5 2 2.5 3 x 10 7 0 10 20 30 40 50 60 70 80 90 100 Frequency Voerde, 10m Antenna Distance, Converted from H-Field using 377Ohm k-factor (dBuV/m - dBm) 80% k-factor med k-factor 20% k-factor Figure 20: k-Factor Measured with Loop Antenna at 10 m Distance and Converted to E-Field An influence in earthing variations (e.g. described in [i.5] clause 6.1) or operator differences was not found when analyzing the EMI results. |
2e7ad6daaa689fc2fb297c1ada7a052a | 101 562-2 | 7.2 Interference Threshold of FM Radio Broadcasts | The level of interference in FM radio broadcasts was tested at 10 different locations. In total, 1,179 subjective evaluations of PLT noise, which can be detected by human ears, in FM radio were conducted. This includes 9 feeding styles times 131 radio stations with various radio devices in a range of different positions. An example of a measurement protocol produced at one location can be seen in table 7. Several radio stations were disturbed with noise signal feeds in various styles through one or more outlets. Various radio devices, battery driven and plugged into a power supply, were monitored. The threshold at which human ears are able to detect interference is noted in the protocol below. Usually, sensitive radio stations, which are difficult to receive, are affected by even the lowest PLT levels. Table 7: Measurement Protocol of FM Interference Threshold in a Home Threshold Frequency Radio station Receiver Injection PNNT EPNT NENT PN EP NE APN PNE in MHz location Point in dBm (120 kHz) into the plug 92,2 SWR 3 Bose System P42 -4,5 -23,5 -31,5 -4,5 -30,5 -19,5 -30,5 -3,5 107,7 Die neue 107,7 Bose System P42 -31,5 -43,5 -39,5 -25,5 -43,5 -40,5 -37,5 -25,5 101,3 Antenne1 Sony Radio P42 1,5 -6,5 -4,5 0,5 -4,5 -5,5 -17,5 -7,5 92,2 SWR 3 Sony Radio P42 -6,5 -19,5 -16,5 -3,5 -18,5 -10,5 -9,5 -0,5 107,7 Die neue 107,7 Sony Radio P42 -22,5 -33,5 -33,5 -33,5 -28,5 -24,5 -29,5 -28,5 101,3 Antenne1 Sony Radio Bat P42 4,5 -4,5 -1,5 5,5 -3,5 -0,5 4,5 5,5 92,2 SWR 3 Sony Radio Bat P42 -4,5 -16,5 -13,5 1,5 -16,5 -16,5 -3,5 1,5 107,7 Die neue 107,7 Sony Radio Bat P42 -18,5 -20,5 -18,5 -13,5 -16,5 -17,5 -27,5 -22,5 107,7 Die neue 107,7 Bose System P42 1,5 -10,5 -5,5 4,5 -9,5 -7,5 -9,5 -8,5 ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 33 -130 -120 -110 -100 -90 -80 -70 -60 -50 -40 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 feeding level into outlet [dBm/Hz] CDF FM interference threshold PNNT EPNT NENT PN EP NE APN PNE Figure 21: Cumulative Distribution Function of Threshold when Interference is Noticeable by Human Ears Figure 21 shows the cumulative distribution of the threshold of all radio services at all locations independent of the radio device or power supply used. The individual lines represent the individual feeding styles. The x-axis is the feeding PSD injected to a power outlet. Table 8 lists the values of the 80 %, 90 %, 99 % and 100 % threshold of when FM radio becomes disturbed. Table 8: CDF Values of Interference Threshold for FM Radio 50 % value in dBm/Hz 80 % value in dBm/Hz 90 % value in dBm/Hz 99 % value in dBm/Hz 100 % value in dBm/Hz PNNT -78 -93 -105 -114 -117 EPNT -81 -95 -103 -114 -115 NENT -81 -96 -105 -114 -125 PN -77 -91 -104 -111 -113 EP -82 -94 -102 -115 -125 NE -79 -92 -102 -116 -119 APN -79 -94 -101 -115 -116 PNE -78 -93 -106 -114 -115 Figure 22 and figure 23 show the influence on a particular radio device (Sony® ICF-SW1000T, see note) when the device is battery driven or power is supplied from the mains. The location of the radio device during both tests was identical. If the radio receiver is battery powered, there is only a radiated coupling path from the mains to receiver. When the radio receiver is mains powered, radiation as well as conductive coupling paths exist. NOTE: Sony® ICF-SW1000T is an example of a suitable product available commercially. This information is given for the convenience of users of the present document and does not constitute an endorsement by ETSI of this product. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 34 -130 -120 -110 -100 -90 -80 -70 -60 -50 -40 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 feeding level into outlet [dBm/Hz] CDF FM interference threshold PNNT EPNT NENT PN EP NE APN PNE Figure 22: CDF of Interference threshold for Sony® ICF-SW1000T with Power Supply ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 35 -130 -120 -110 -100 -90 -80 -70 -60 -50 -40 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 feeding level into outlet [dBm/Hz] CDF FM interference threshold PNNT EPNT NENT PN EP NE APN PNE Figure 23: CDF of Interference Threshold for Sony® ICF-SW1000T Battery Driven The graphs in figure 22 and figure 23 are pretty much identical. The conductive coupling path from mains interference to the radio device seems not to be dominant in the VHF range. Or in other words, the FM radio used in this test is sufficiently isolated from the mains interferer. This phenomenon is different compared to the HF range [i.3]. The STF 410 FM radio immunity field tests were performed using the Sony® ICF-SW1000T device and the radio device was owned by the habitant of the flat. The immunity of all radio devices was pretty much identical. No dependency among device manufacturers or HIFI radio systems versus kitchen radios could be found. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 36 Annex A: Alternative Procedure for NWA Calibration if Amplifier Output Power is too high for NWA Input If the output power of the amplifier is too high for the NWA input, an attenuator (of sufficient power) can be inserted between amplifier output and NWA input during calibration (see figure A.1). An attenuation of 30 dB is recommended. NWA amplifier cable used for signal injection attenuator antenna cable (with ferrites for suppression of sheat current) s‘21 Figure A.1: Alternative NWA Calibration The "through" calibration is made in the usual way. During measurements, the attenuator is removed. The setup is shown in figure 5. The operator has to correct the measured data before archiving them in the STF 410 database. The true s21 value is derived from the measured s21 values by subtracting the attenuation of the attenuator. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 37 Annex B: Software for Automatic File Naming B.1 General According to experience collected from other measurement campaigns, a structured procedure for acquiring the wanted data is sometimes critical. In [i.9], clause A.2 a scheme based on a page that should be completed during tests is used for this structured approach. Here software has been developed in order to avoid different file naming, typographical errors in file naming and similar errors, which at the end would lead to manual corrections for the statistical evaluation. Although the software is equipped with an interface to readout the NWA and store the data in the correct file, it can also be used for managing just the measurements. All data is stored in an ASCII-form so it can be read with a large number of software tools for further investigations. The software is available to all STF 410 members as a resource. B.2 Main Dialog Main dialog box after opening the software. Figure B.1: STF 410 Software, Site Description The first step is to assign an appropriate file folder for the site specific data and measurement results. The folder has to be created in advance. After selecting the Menu "Workspace" a dialog box will show up, with which the folder can be assigned. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 38 After that, the user should fill in the general information associated with the measurement site, i.e. Country, City, Zip, Street, Building No (if applicable), GPS Coordinates, Date and Time of the Measurements, Persons, who performed the measurements. The general descriptions can be completed with arbitrary comments. To allow better interpretation of results obtained from the specific measurement site, some photos including remarks on what is shown on the photos are helpful. This information is stored with "Save". After the general information, the associated antenna locations and feed locations (plugs) need to be defined. This is easily be done by placing a unique ID for the location in the text box below the tables for feed files and antenna locations. A click on the "Add" button will put the ID into the corresponding list. B.3 Antenna Location Description Dialog A pre-defined antenna location can be edited by double clicking on an entry in the antenna locations table. The following dialog appears. Figure B.2: STF 410 Software, Antenna Location Again some general information can be given to describe this special antenna location. Also some photos showing the antenna and its relation to the building are helpful. Loop antenna or biconical antenna are the possible antenna types. The antenna height and distance information is needed for an automatic evaluation and statistics. When "Set Filenames" is pressed the text boxes in the frame Noise are filled. Since these files are not yet present, the background is set to a red color. If the files exist it will be green. If a recognized measurement receiver is connected to the PC a click on the button "Measure" will perform the measurement and store the data in the selected file. In the version on the FTP-site the receiver ESIB from Rohde&Schwarz is supported. Support of other receivers may be available upon request. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 39 B.4 Feed Point Description Dialog A pre-defined feed file can be edited by double clicking on an entry in the antenna locations table. The following dialog appears. Figure B.3: STF 410 Software, Feeding Settings Again some general information can be given to describe this special feed location. Also some photos showing the plug with the attached adapter and its relation to the structure of the building are helpful. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 40 The injection of signals to the mains causes variations in field strength at certain antenna locations. This can be documented by selecting one of the predefined antenna locations in the software. A press of the "Set filenames" button defines all the necessary file names and measurements. If a NWA is connected to a PC, the measurements can be directly started by selecting a certain permutation and pressing the "Measure" button. NWA ZVRE from Rohde & Schwarz is supported by the version on the FTP-site. Support of other NWA may be available upon request. The file is marked green upon completion to indicate that the measurement has been done. Pressing the "View" button calls the graphic tool Gnuplot (public domain tool) for graphical representation of the measured data. Besides the pure EMI-measurement the software also manages Noise, S11 and Transfer Function measurements by selecting the corresponding tabulator. B.5 Help for Injection Types A help function can be called from the main dialog (menu entry Help - Coupling Types) or by double clicking on the coupling type (e.g. P-E) in the feed file dialog to assist with setting switches of the box correctly. Figure B.4: STF 410 Software, Coupling Types This dialog shows how to adjust the switches, which BNC plugs will be terminated and how the measurement instruments are connected. The different permutations can be shown in this dialog simply by clicking on the coupling type on the left side. B.6 File Formats For better exchange a simple ASCII-Format is used for data storage. For data measured with the network analyzer, there should be one header line indicating the number of points and some settings, followed by a table with the measurement results. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 41 Its form looks like: Number of points: 1601, Startfrequency: 1e+006, Stopfrequency: 1e+008, Label:'XFR:POW:S12' 1e+006 -0.000272393 -0.000801793 0.0008468 -1.89829 -61.4444 1.06188e+006 -0.000102707 -0.000721605 0.000728877 -1.71218 -62.7469 1.12375e+006 -0.000300805 -0.000974474 0.00101985 -1.8702 -59.8293 ... The values in the table are: frequency in Hz, s12 real part, s12 imaginary part, s12 magnitude (linear), s12 phase, s12 magnitude (in dB) For data from a measurement receiver there is no header. The data is just stored as a table, which looks like this: 1000000 23.5484161376953 9.04364013671875 1005000 21.0335845947266 14.4008636474609 1010000 22.5505752563477 14.3177490234375 ... The values in the table are: • frequency in Hz; • peak detector reading in dB(µV); • average detector reading in dB(µV). B.7 Creation of Data for the FTP Server Since the software uses an internal file naming scheme, the file names and file formats need to be converted into the format defined in clause 6.3.1. This can be done by selecting "rename files to WI28 and WI29" from the "Tools" menu of the main dialog. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 42 Annex C: Bibliography Terms of Reference for Specialist Task Force STF 410 (TC PLT) on "Measurements to Verify Feasibility of MIMO PLT", version: 1.1, 6 May 2010. ETSI ETSI TR 101 562-2 V1.3.1 (2012-10) 43 History Document history V1.1.1 August 2011 Publication V1.2.1 February 2012 Publication V1.3.1 October 2012 Publication |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 1 Scope | The present document addresses location systems combining telecommunication networks with Global Navigation Satellite System (GNSS) and other navigation technologies in order to deliver location based services. The analysis contained in the present document is intended to highlight the growing use of complex location systems in order to deal with the expansion of location based applications in a mass market. The objective is thus to demonstrate both relevancy and achievability of standardising a high-level architecture for these systems, and the associated minimum performance. In order to achieve this objective, the present document first provides a reminder of the types of applications which rely on location information provided by such systems in order to provide services. Secondly, it describes these location systems, in terms of key functions to be fulfilled (also called key features) and available enabling technologies at system components level (navigation sensors, hybridization techniques). It also focuses on the definition of operation environments applicable to such systems (depending on the type of application). Finally, preliminary location systems architecture, interfaces and performances are defined. It is highlighted that the scope of this technical work specifically excludes standardisation of safety of life applications related to civil aviation, which are already addressed through, in particular, Radio Technical Commission for Aeronautics (RTCA) standards. |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 2 References | References are either specific (identified by date of publication and/or edition number or version number) or non-specific. For specific references, only the cited version applies. For non-specific references, the latest version of the reference document (including any amendments) applies. Referenced documents which are not found to be publicly available in the expected location might be found at http://docbox.etsi.org/Reference. NOTE: While any hyperlinks included in this clause were valid at the time of publication, ETSI cannot guarantee their long term validity. |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 2.1 Normative references | The following referenced documents are necessary for the application of the present document. Not applicable. |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 2.2 Informative references | The following referenced documents are not necessary for the application of the present document but they assist the user with regard to a particular subject area. [i.1] ETSI TR 103 183 (V1.1.1): "Satellite Earth Stations and Systems (SES); Global Navigation Satellite Systems (GNSS) based applications and standardisation needs". [i.2] RTCA DO-229D (2006-12): "Minimum Operational Performance Standards for Global Positioning System/Wide Area Augmentation System Airborne Equipment". [i.3] OMA-TS-ULP-V2-0-20100816-C (2010-08): "User Plane Location Protocol". [i.4] ETSI TS 122 071 (V9.0.0): "Digital cellular telecommunications system (Phase 2+); Universal Mobile Telecommunications System (UMTS); LTE; Location Services (LCS); Service description; Stage 1 (3GPP TS 22.071 version 9.0.0 Release 9)". ETSI ETSI TR 101 593 V1.1.1 (2012-09) 7 |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 3 Definitions, symbols and abbreviations | |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 3.1 Definitions | For the purposes of the present document, the following terms and definitions apply: application central part: entity hosting the Location based application, that interacts with a Location system central facility in order to obtain Location information for one or more mobile targets NOTE: The application central part can be located inside or outside the Positioning terminal. location based application: application which is able to deliver a service to one or several users, built on the processing of the Location information related to one or several mobile targets location information: information about the position and other related information of a positioning terminal NOTE: It is the main output of a Location system. The information can be of several types: not only the terminal position itself, but also the reliability of the reported position, the authenticity of the reported position, the probability that the terminal is/was in a given pre-defined area, information related to the terminal motion (speed, acceleration, gyros, etc.), etc. location system: infra-structure for reporting to a location based application the Location information of one or several positioning terminals, periodically or upon request location system central facility: logical entity, inside a Location system, that manages the location information exchange protocol with the application central part, which is the location system external client mobile target: physical entity whose position the location system builds the location information on, and to which the positioning terminal is attached positioning terminal: logical entity, inside a Location system, that contains a GNSS receiver and possibly additional sensors, and is physically located with the mobile target NOTE: It executes the measurements needed to determine its position, and implements part of the location determination functions. It embeds the group of sensors needed to execute these tasks. This group can include navigation sensors (GNSS, Inertial, Odometers, etc.), wireless network modems (terrestrial or satellite). |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 3.2 Symbols | For the purposes of the present document, the following symbols apply: ϕ Carrier phase δAccel Error on sensor acceleration (from INS) δAtt Error on sensor attitude (from INS) δGyro Error on sensor gyroscopes (from INS) δPos Error on sensor position (from INS) δPos3D Uncertainty on sensor position (from GNSS) δV Error on sensor attitude (from INS) δV3D Uncertainty on sensor speed (from GNSS) d Carrier Doppler PGNSS Position estimate coming from GNSS sensor PINS Position estimate coming from the INS VGNSS Speed estimate coming from GNSS sensor VINS Speed estimate coming from the INS ETSI ETSI TR 101 593 V1.1.1 (2012-09) 8 |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 3.3 Abbreviations | For the purposes of the present document, the following abbreviations apply: 3GPP 3rd Generation Partnership Project ADAS Advanced Driver Assistance Systems AL Alarm Limit BTS Base station Transceiver System DOA Direction Of Arrival ECEF Earth Centred Earth Fixed EDGE Enhanced Data for GSM Evolution EGNOS European Geostationary Navigation Overlay System EMI Electro-Magnetic Interference FDAF Frequency Domain Adaptive Filtering GCF Global Certification Forum GEO Geostationary Earth Orbit GIVE Grid Ionospheric Vertical Error GLONASS Global Navigation Satellite System (Russian based system) GNSS Global Navigation Satellite System GPRS General Packet Radio Service GPS Global Positioning System GSM Global System for Mobile communications HPE Horizontal Positioning Error HPL Horizontal Protection Level IMU Inertial Measurement Unit INS Inertial Navigation Sensor IRS Inertial Reference System ITS Intelligent Transport Systems LCS LoCation Services LEO Low Earth Orbit LOS Line Of Sight LTE Long Term Evolution MEMS Micro Electro-Mechanical Systems MEXSAT Mexican Satellite System MI Mis-Integrity MMI Man-Machine Interface MOPS Minimum Operational Performance Specification MP Multipath MPS Minimum Performance Standard MS Mobile Station NCO Numerically Controlled Oscillator NMR Network Measurement Results ODTS Orbit Determination and Time Synchronisation OMA Open Mobile Alliance OTDOA Observed Time Difference Of Arrival PAYD Pay As You Drive PE Positioning Error PL Protection Level PRS Public Regulated Services PVT Position, Velocity and Time QoS Quality of Service QZSS Quasi-Zenith Satellite System RAIM Receiver Autonomous Integrity Monitoring RF Radio Frequency RMS Root Mean Square RTCA Radio Technical Commission for Aeronautics RTK Real Time Kinematic SBAS Satellite Based Augmentation System SCN Satellite Communications and Navigation (Working Group of TC-SES) SMLC Serving Mobile Location Center SUPL Secure User Plane for Location ETSI ETSI TR 101 593 V1.1.1 (2012-09) 9 SV Satellite Vehicle TBC To Be Confirmed TBD To Be Defined TC-SES Technical Committee Satellite Earth Stations and Systems TTA Time To Alarm TTFF Time To First Fix UDRE User Differential Range Error UERE User Equivalent Range Error UHF Ultra-High Frequency UMTS Universal Mobile Telecommunications System VPL Vertical Protection Level WAAS Wide Area Augmentation System WI-FI Wireless Fidelity |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 4 Location based applications needs | A thorough inventory of the location based applications is done in TR 103 183 [i.1]. This inventory is used as a reference in the present document. The present clause only acts as a reminder of the conclusion of this inventory, presented as a classification of these applications. It is indeed possible to organize applications facing similar needs in a reduced number of classes: a) Location Based Charging: - The objective is to charge a user based on its reported position. The main requirements are: Reliability of check point crossing detection: there is a risk that user reported position triggers a charging event whereas it is actually in a position free of charge. This risk is generally needed to be very low. The service availability: the percentage of cases when user actual position has to trigger charging event, but system is not properly informed. The service unavailability can be due either to an erroneous reported position, or the unavailability of the location information itself. This service unavailability is generally needed to be low. NOTE 1: This type of location-related requirement is needed for road user charging (road), on-street parking fee pricing (road), waterways and harbours charging (maritime/multimodal), home zone billing, regulated fleets in urban areas, etc. b) "Pay As You Drive" (PAYD) charging: - The objective is to charge a user based on the travelled distance (mainly applicable for pay-per-use insurance). The challenge is quite similar to the previous group, except that useful information is rather the travelled distance than the position itself. - The main drivers are: the representativity of the computed distance; or the representativity of the followed trajectory; in order to globally optimise the fee collection. NOTE 2: This type of location-related requirement is needed for pay per use insurance (road), car rental pricing (road), taxi service pricing (road), freight tolling (road), car pooling (road), pay as you pollute (road), energy charging (train). ETSI ETSI TR 101 593 V1.1.1 (2012-09) 10 c) Cooperative basic geo-localization (including fleet and asset management): - The objective is to recover the position of one or several assets or vehicle, remotely or otherwise. The main drivers are generally: The reported position accuracy: as far as fleet management or personal navigation is concerned, the main target is explicitly to obtain an accurate position estimate. The required accuracy highly depends on the application: tens of meters for personal road navigation and vehicle fleet management, meters for pedestrian personal navigation and city sightseeing. The service availability: position availability might not be as driving as for other application (see location based charging applications), but it is a clear challenge in the considered applications: car positioning in urban area (including high masking or shadowing, tunnel, important multipath) clearly suffers from degraded availability. NOTE 3: This type of location-related requirement is needed for fleet/asset/resource management, personal navigation (pedestrian, road, multi-modal), traffic travel info, city sightseeing, etc. d) Non-cooperative geo-localization (possibly applied to fleets): - Asset positioning might be required when asset is non-cooperative. In other words, compared to the regular "cooperative basic geo-localization", a new driver is reported: the service reliability. - In other words, this new requirement is important any time the terminal is placed in a "hostile" environment, and that the confidence in the reported position is maximized. NOTE 4: This type of location-related requirement is needed for some kind of fleet management (car rental), car recovery after theft (road), city logistics (road). e) Reliable geo-localization (including dangerous, precious and/or sensitive cargos): - The objective is to obtain a reliable position estimate for any application where position is a key driver for security or safety (of cargo, travellers). - The main driver here is the confidence level associated to the applicative figure of merit. This figure of merit can be: the reported position; application event: billing event, trajectory. - In other words, for such applications it becomes paramount to be informed of the probability that reported information is inaccurate. - Of course, reported position accuracy and service availability are important drivers, which might however depend on the specific applications. - The border between "non-cooperative" and "reliable" geo-localization is thin. To that point, they are however considered separately: Non-cooperative geo-localization only targets position uncertainty caused by position spoofing (i.e. wanted). In other words, any position uncertainty due unwanted origins (GNSS signal, interference, other) are not covered: they are deemed naturally bounded, and the application required accuracy is compatible with this bound. Reliable geo-localization however covers all sources of position uncertainty, in order to bring confidence not only in the position authenticity, but position accuracy. NOTE 5: This type of location-related requirement is needed for livestock transport tracking and tracing survey, dangerous and hazardous cargoes tracking and tracing survey, special (high value, sensitive, dual) goods traffic tracking, perishable goods / food tracking and tracing. ETSI ETSI TR 101 593 V1.1.1 (2012-09) 11 f) (Reliable) Vehicle movement sensing: - Some application aim at collecting, in addition to the terminal position, additional information related to its movement: speed, acceleration, heading, gyration, etc.: The main driver is of course the movement caption accuracy. The objective might to measure vehicle speed for law enforcement of eco-driving advice. As previously mentioned, a confidence level associated to the reported parameter might also be needed. NOTE 6: This type of location-related requirement is needed for: Liability critical applications: legal speed enforcement (road), accident reconstruction (road), vehicle control assistance (ADAS) + collision warning (road), cold movement detector (train), traffic management systems (train). Non-liability critical applications: eco-driving and carbon emissions foot- printing (road), traffic congestion reporting (road). |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 5 Location systems | In the previous clause, a set of location systems functional needs has been established. Clause 5 now proposes a description of the location systems allowing to achieve the identified function. The first step is to provide a generic context in which location systems are likely to step in. Secondly, we intend to define a high level technical specification applicable to location systems, in order to define location systems key features allowing to fulfil the required functions. Then, a review of the enabling technologies allowing to support these key features is executed. Finally, the operational environments in which the location systems are expected to perform are also reviewed. Consequently, we first propose to review generic architecture of a location system, and then list the location system key features that need to be addressed in order to answer to the application level needs. |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 5.1 Generic context | In order to better capture the purpose of location systems, this clause provides a description of the generic context in which Location systems are needed. Figure 1 thus shows how such systems could be integrated to other systems and external service providers in order support location based service. The considered scenario envisions a location based application, which delivers to an external entity (unknown, out of the present work) a service based on the location of a number of mobile targets. The central part of this application thus addresses a request to a location system. ETSI ETSI TR 101 593 V1.1.1 (2012-09) 12 Figure 1: Generic context for location systems use In the situation described above: • The GNSS receiver inside the positioning terminal offers the ability to track the signals received from the system(s) constellation(s), and thus provide measurements or position estimate. • Additional sensors are also available inside the terminal, in order to provide complementary location-related data. • Both GNSS receiver and sensor are integrated in a single device, identified as a positioning terminal. • Whenever a location request, initiated by the application central part (external location system client), is processed and transmitted by the Location system central part to the positioning terminal, it triggers measurements at terminal level, resulting in location data (either terminal position or specific measurement). • This data sent back to the location system central part for further processing, consolidation and formatting. • Furthermore, an interface between the ground infrastructure of the GNSS or SBAS systems and the location system central part can also be considered, through a dedicated communication mean. This interface allows to convey GNSS navigation data as uplinked towards the spacecrafts, or augmentation data. The communication link used is possibly different from the one connecting the positioning terminals to the central part. However, in figure 1, the same "communication system" box gathers both means of communications. The generic architecture given above is a high level description of location systems use. It provides the generic framework, which adapts to most of the conceivable use cases (see example in clause 4). |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 5.1.1 Architecture example from existing systems | Historically, it is considered that GNSS is the primary means of location. But it has to be complemented by other location means to satisfy the location service availability requested by the some application and users. A very good example is inside the 3GPP framework. In this case, the Location system is made of the Mobile Station (MS), equivalent to the positioning terminal, the SMLC (Serving Mobile Location Center), equivalent to the Central facility, and it includes a communication function that relies on the mobile communication network. This is fully described in TS 122 071 [i.4]. The following is intended to show how such location systems operate. It also illustrates how the positioning terminal is embedded in such architecture. ETSI ETSI TR 101 593 V1.1.1 (2012-09) 13 GNSS systems infrastructure are basically composed of the following segments: • The space segment, composed of the satellites constellation that disseminate the signal- in- space to the user segment. • The ground segment, including the ground control segment devoted to the control of the satellites constellation, and the ground mission segment devoted to the preparation of all the data to be included within the signal -in -space. • The user segment, including any GNSS receiver that uses the GNSS signal- in- space to determine its position. Other sensors may exist in the terminal to enhance the positioning performance in difficult environments. The main GNSS today operating or under development are GPS, Galileo, Glonass, modernized GPS, and Compass. In addition to GNSS, Satellite Based Augmentation System also allows to support wide-area or regional augmentation through the use of additional satellite-broadcast messages. Such systems are commonly composed of multiple ground stations, located at accurately-surveyed points. The ground stations take measurements of one or more of the GNSS satellites, the satellite signals, or other environmental factors which may impact the signal received by the users. Using these measurements, information messages are created and sent to one or more satellites for broadcast to the end users. Galileo studies also considered local elements that are NOT part of the GNSS infrastructure, but bring further location data to the GNSS receiver to allow them to locate themselves in certain difficult conditions. These local elements may be for example, mobile communication networks elements (GSM Base station Transceiver System (BTS), UMTS "NodeB"), that provide RF signal used by positioning terminals, or dedicated equipment such as pseudolites. An interface appears between the GNSS ground mission segment, and the mobile communication network location infrastructure, to convey the GNSS navigation data that may be acquired from the signal- in- space using a GNSS receiver, but that may also be disseminated by the GNSS ground mission segment using a terrestrial communication link. The following figures summarize the location provisionning systems, introducing also the concept of positioning (or location) terminal, which embeds a GNSS receiver, and additional location sensors. The exchange of information are sequentially numbered on the arrows between the entities. The proposed example corresponds to the case where the location request is initiated by an application running on an external network (network initiated case). ETSI ETSI TR 101 593 V1.1.1 (2012-09) 14 Location terminal SMLC (Control Plane) SPC (User Plane) + A-GNSS Server Application Server Local Elements GMLC(Control Plane) Or SLC (User Plane) GNSS Space Segment (constellations) Mobile Communication Network 0 Signal In Space 4 Positionning data 1 Location request 2 Authorized Location request 3 Location Transaction 5 Location response 6 Location response GNSS Ground Segment (Mission) 0 GNSS Navigation data via Cround communication Link GNSS Mission data Mobile Communication Network Location infrastructure Figure 2: Example of location system in the 3GPP context Mobile communication network location infrastructure elements are fully defined in the 3GPP standards dedicated to location, and also in Open Mobile Alliance (OMA) standards dedicated to location using the Secure User Plane for Location (SUPL) protocol. Functionally speaking, location servers in both domains are equivalent. For more information concerning the functions borne by the mobile communication network location infrastructure elements, please refer to TS 122 071 [i.4]. |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 5.2 Location systems key features | Based on the inventory of location based applications, it is now proposed to derive the identified key functions into location system technical specification. "Technical specification" should be understood here as follows: functions previously listed are expressed in terms of application user requirements. Derivation of associated location system technical specification consists in translating them into technical requirements, exploitable at location system level. The main technical requirements derivated are the key features presented in this clause. As far as location needs are concerned, a list has been established in clause 4. Table 1 summarizes these needs, and proposes a list of technical key features allowing to fulfil these needs and a mapping to these needs towards the location system key features. ETSI ETSI TR 101 593 V1.1.1 (2012-09) 15 Table 1: Application needs and location systems key features Mapping: Application needs vs Key features Location system key features position horizontal accuracy Position authentication Position availability Position integrity Interference rejection Application needs Reliability of the detection of check point crossing (location based charging) X X X The billing service unavailability (location based charging) X X Representativity of the computed distance (PAYD) X X X Representativity of the reported trajectory (PAYD) X X X The reported position accuracy (cooperative and non- cooperative geo-localization, reliable geo-localization) X The location service availability (cooperative and non- cooperative geo-localization, reliable geo-localization) X Service reliability (non-cooperative geo-localization) X X X Confidence level associated to the applicative figure of merit (reliable geo-localization) X X Movement caption accuracy (vehicle movement sensing) X X Confidence level associated to the reported parameter (reliable vehicle movement sensing) X X In order to further describe the proposed key features, and support the above mapping, the following clauses describe these features. |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 5.2.1 Position horizontal accuracy | |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 5.2.1.1 Feature description | This is the most common location system performance, and is quite explicit: accuracy is measured as the error between the reported position and the actual position of the sensor. Horizontal accuracy is the derived 2D position error. NOTE: Vertical accuracy could also be taken on-board the technical framework, but none of the considered applications explicitly require for position vertical accuracy. |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 5.2.1.2 Feature metric | It is usually expressed in meters, as the Root Mean Square (RMS) of the Horizontal Positioning Error (HPE). The horizontal positioning error is computed as the distance between the 2D reported position and the 2D true position (i.e. projected on the horizontal plane). |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 5.2.1.3 Example of implementation | The development of location sensor able to provide accurate position estimate is the basic challenge taken by any GNSS receiver manufacturer, with comparable performances for each of them. ETSI ETSI TR 101 593 V1.1.1 (2012-09) 16 The techniques behind these implementations being very similar among them (at least the basic ones) and well documented in many GNSS bibliographic sources, no details are provided here. |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 5.2.2 Position authentication | |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 5.2.2.1 Feature description | As exposed previously, a significant number of applications are sensitive to the reliability of the reported/computed position: "Can I trust the position reported by such asset?". At the location system level, this is translated into a need for position authentication. The concept of authentication is proposed to be analysed in this clause. GNSS service vulnerabilities: Geo-localization using GNSS signals relies on the processing of the input Radio Frequency (RF) signals coming the satellites system. Among all available systems, the signal most often offered are unencrypted signals (GPS L1C/A, GPS L1C, GALILEO Open Service, GLONASS, etc.). These signals are clearly GNSS system weakpoint, since it is theoretically and practically very "easy" to spoof a receiver with a RF signal generator, replaying signals recorded at another location: if appropriately modified, software receivers can be converted into spoofers by reverting the receiving chain, adding some offsets to each satellite signal and irradiating a modified version of the received signal in the air. The objective is clear: in a location based charging application, a non-cooperative user or attacking third party can: • Record the GNSS signals when using a toll free road (red path below). • While using the paying road (blue path below), replay the recorded RF signals, and thus report to the monitoring system a spoofed trajectory, corresponding to a toll free use case. Figure 3: Example of spoofed trajectory versus true trajectory The motivation is clear: avoid billing (money), law enforcement (prisoner monitoring), spoof a positioning unit to hijack a monitored cargo or asset. ETSI ETSI TR 101 593 V1.1.1 (2012-09) 17 Authentication solutions: Three main threads are explored to implement authentication: 1) Technological thread: - It mostly concerns the possibility to hybridize regular GNSS receivers with additional sensors, into a positioning terminal. The key methods identified relied on hybridisation again following two axes: Hybridisation with low cost inertial, magnetic sensors, odometers. Hybridisation with network positioning (namely cell-id and Network Measurement Results (NMR)). NOTE: Such a function can be fully implemented on-board the device, which turns into a positioning terminal rather that a standalone GNSS receiver. 2) Profiling thread: - This centralises all the possible improvement brought to the authentication performance through the use of the user's dynamic model. It can be derived either from known mechanical characteristics, or from known use cases. - Such function is usually implemented at location system infrastructure level, to which the terminal is connected through the communication network, since access to user data base is needed. 3) Encrypted context thread: - Authentication can also be achieved exploiting military/regulated services encryption feature, but on a regular (i.e. non-military) location terminal. The basic principle is: GNSS receiver can sample a short period of RF signal. Transmit it to a location center imbedding the capability to correlate with the military/regulated spreading codes. and therefore ensure that the RF signal fed into the GNSS receiver is indeed collected from the regular constellation. |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 5.2.2.2 Feature metric | The authentication performance is proposed to be measured as the fraud detection performance. In other words, two figures of merit are used: • the mis-detection performance; • the false-alarm rate. NOTE: The performance depends on the nature of the threat. This is addressed in clause 6.3.2 Authentication. |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 5.2.2.3 Example of implementation | Using the above techniques, simulations have been executed. The scenario is: • the mobile is equipped with GNSS receiver and an Intertial Navigation Sensor (INS); • it follows a given "true" trajectory; • it is spoofed with a GPS signal previously recorded, while following another trajectory; • the discriminating criteria used to detect spoofing (and build authentication) are the heading and the heading change rate information. ETSI ETSI TR 101 593 V1.1.1 (2012-09) 18 Figure 4: Authentication means description We can clearly see that: • heading recovered from un-spoofed GPS location (green curve) and INS (dark blue curve) are consistent; • heading computed from the spoofed GPS location (red curve) is totally inconsistent; • based on the comparison of this information, flags can be raised any time the consistency of information coming from the various sensors drops below a certain level (light blue curve for heading info, purple curve for heading change rate information, and dark curve for the spoofing flag obtained via consolidation of the two previous ones). Similar concept can be applied to communication sensors data, which also provide a location estimate (even if it is a coarse one): • along a given trajectory, terminal connects successively to different Base Station Transceiver Systems (BTS); • each BTS as a known position and coverage area; • a spoofing attempt can lead to inconsistent information between reported GNSS position and serving cell identification; • the longer the trip is, the more likely is the spoofing detection. ETSI ETSI TR 101 593 V1.1.1 (2012-09) 19 Base Station Base Station Base Station (BTS) Emitteur-Receiver Base Station Controle Coverage of the cell Positionning Node Figure 5: Authentication using serving cell IDs along a terminal trajectory |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 5.2.3 Position availability | |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 5.2.3.1 Feature description | Position availability is a key driver, in particular for multi-modal applications (outdoor to indoor). It represents the availability rate of the position information at location system output. |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 5.2.3.2 Feature metric | It is expressed an availability rate, usually expressed as a percentage. |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 5.2.3.3 Example of implementation | For liability-critical applications, availability performance is inter-correlated with integrity. This is developed in the next clause. |
d397fedfd9b02ae777b5d247c5d97147 | 101 593 | 5.2.4 Position integrity |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.