hash stringlengths 32 32 | doc_id stringlengths 7 13 | section stringlengths 3 121 | content stringlengths 0 2.2M |
|---|---|---|---|
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 6.6.0 Introduction | The RCST builds a set of CRs for each RC using DA. This is achieved through the coordination of HL and LL Services. LL Services provide mechanisms to access satellite resources, such as RA or DAMA. HL Services may be used to implement IP services, such as traffic queuing and support for PHBs and are mapped to LL Services. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 78 The characteristics of the available RCs should be considered when selecting a mapping for a HL Service to realize a DiffServ PHB. An HL Service implementing the EF PHB requires a bandwidth guarantee and is typically used to provide low loss, low latency and low jitter service. EF traffic tends to vary slowly. CRA is a suitable option to meet these requirements. RBDC may be used to supplement the CRA, but it may not be desirable to use this alone because the delay to obtain an RBDC assignment may not meet the low latency requirement. The rationale behind of using RBDC to supplement CRA is its offer a higher efficiency: users may tolerate the increase of delay and delay variation due to RBDC in exchange for a more efficient allocation. Since EF traffic requires low jitter, the HL Service may assign this traffic to a higher scheduling priority and use a SA that allows pre-emption of jitter-tolerant traffic. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 6.6.1 Example of support for basic DiffServ QoS | An HL Service that implements the AF PHB demands a delivery guarantee but does not mandate constraints on delay or jitter. Since AF traffic may be highly variable, dynamic bandwidth allocation methods are recommended. An RC that supports RBDC may be suitable to meet the AF requirements, since it offers a bandwidth guarantee (up to a certain level) and can efficiently handle bursty traffic. VBDC may also be used to supplement the capacity requested via RBDC. Although VBDC may be assigned with lower priority in some systems, it can respond to a request with high efficiency. Use of VBDC may allow a trade-off between offered QoS and the cost of capacity. Since AF traffic does not require low jitter, the HL Service may assign this traffic to a scheduling priority that is lower than EF and use a SA that allows pre-emption of this traffic by jitter-sensitive traffic. An RCST may implement several HL Services offering AF support, or variants of the HL Service adapted to the needs of specific classes of traffic. The variants may be mapped to the same SA. An HL Service that implements the Best Effort PHB has no specified requirements. Consequently, BE traffic may be mapped to an RC that uses VBDC only. Scheduling should ensure that this class does not take resources away from EF or AF services, and traffic should not share the same SA as used for EF. An RC may be mapped to multiple Capacity Categories, in which case a weight may be applied to the different categories. Different studies have been performed to reach the best way to map capacity categories into PHBs. Some example request strategies are proposed below, based on periodic assignments: EF = 80 % CRA + 20 % RBDC AF = 70 % RBDC + 30 % VBDC BE = 100 % VBDC Using the above mapping, an RBDC rate may be requested that is equal to 20 % of the ingress rate in the EF queue + 70 % of the new ingress rate in the AF queue. Possible mappings between allocation channels and SAs are: • A mapping that assigns all allocated timeslots to a single SA. • A mapping that provides strict separation between a set of SAs. Assuming the first type of mapping, the RCST will use the HL service information to identify the relative priority between each BA (i.e. BA queue priorities) that is mapped to the same SA. As an example, the scheduler could assign the highest priority to PDUs queued in a BA with an EF PHB (to satisfy low latency requirement), followed by the BAs with an AF PHB, and finally the BE traffic, which may receive the lowest priority. Any portion of the total assigned capacity that is not used by a higher-priority queue would be available to a lower-priority BA. This optimizes the use of the according LL service. In this example, a higher-priority queue can hence use capacity allocated in response to a previous CR for a lower-priority queue, allowing it to serve the new incoming higher priority traffic with less delay. However, this creates a temporary condition of under-allocation for the lower-priority queues (observable as increased delay for the BAs assigned to these queues). The queued lower priority traffic will be sent when capacity is allocated in response to the CR for the high priority CR. Traffic conditioning may be used to control the amount of traffic admitted to a BA with the EF PHB to control this "capacity borrowing". ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 79 |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 6.6.2 Example 1 - RCST configuration | The diagram in Figure 6.22 shows an instantiation of the general QoS model. It illustrates the relationships between modules and identifies the HL QoS functions and the LL QoS functions. Solid lines represent the flow of PDUs and other data through the system, whereas dashed lines are used to denote control relationships. Hexagons represent functions and rectangles represent QoS objects. Figure 6.22: Example of QoS model instantiation In this example, IP traffic arriving at the LAN interface of an RCST is routed to a CA. The CA is classified into four BAs according to their DSCP code-point: • EF traffic is mapped to a BA HLS PDU queue, managed according to the EF PHB. • AF traffic is mapped to one of two BA HLS PDU queues, managed according to the AF PHB. • BE traffic is mapped to one BA HLS queue, managed according to the BE PHB. Three HL Services in the control plane are instantiated to handle the set of BAs: EF, AF and BE. The HL Services are mapped to three distinct LL Services: the EF to the Real Time (RT) service, the AF to Jitter Tolerant (JT), and DF traffic to the Best Effort (BE) LL Service. Each LL Service is characterized by its respective RC. Table 6.10 lists recommendations for mapping rate parameters for PHBs to DVB-RCS2 capacity limit values to derive the configuration parameters for the RCs. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 80 Each BA (HLS PDU queue) is mapped to a SA. A scheduler operates on the traffic forming a SA. The scheduler may be triggered each transmission opportunity (notified by the TBTP2) to select the PDUs to be segmented/encapsulated into the LL stream. In this case, a set of SAs are considered, the selection of PDU is based on HL Service parameters and LL Service information. The mapping to a LL Service ensures that PDUs or segmented PDUs are sent using the corresponding AC. When required, PDUs pass through a segmentation function, so that any unsent data is postponed to a later scheduling opportunity. Each segment from an SA is then encapsulated the corresponding configured stream (Str in the diagram) and is then placed in the burst for transmission. This scheduler could, for example, use a strict priority scheduler or a weighted priority algorithm. In this example, three LSs are configured: the RT stream can pre-empt the nRT or BE streams. The example configuration parameters for the HL Services are summarized in Table 6.10. A set of TCs defines the mapping of traffic to the corresponding HL Service. The TC may optionally define conditioning parameters (e.g. in terms of bandwidth or rate). The assumption is that only two rate parameters are used, as a part of an IP flow profile - the Assured Rate and the Peak Rate, the meaning of these parameters is dependent on the HL Service. Table 6.10: Example of QoS bandwidth parameters DVB-RCS2 PHB DVB-RCS2 RC for HL Service RC Capacity Category parameters Comments EF PHB Real Time CRA = ΣAssured rate RBDCMax = 0 VBDCMax = 0 CRA is the guaranteed IP flow rate AFn PHB Critical Data CRA = 0 RBDCMax = γΣGuaranteed rate VBDCMax = γΣ(Peak - Guarantee)rate If there is assured rate, it can be mapped to CRA. RBDCMax is a fraction (γ) of the sum of the guaranteed rate VBDCMax is a fraction (γ) of the sum of difference between the guaranteed and peak rates. BE PHB Best Effort CRA = 0 RBDCMax = 0 VBDCMax = δΣSDR' VBDCMax is a fraction (δ) of the total peak rate. NOTE 1: All parameters apply to the uplink return path, therefore they are transmit parameters. NOTE 2: Assured and Peak Rate are per aggregate rates requested by an RCST. NOTE 3: With the exception of AFn services, all other services map to a single RC. In the case of an AF services, several HL Service instances (n=1-4) are typically mapped to one RC. Therefore, the RC capacity limit values should reflect the aggregation (sum) of individual rates, possibly with an allowance for statistical multiplexing. NOTE 4: The sum of all capacity limit values for all request classes should not exceed the RCST transmission rate. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 6.7 Recommendations for the use of Request Classes | Each RCST is configured to support at least one RC, with some combination of capacity categories. The set of methods is authorized by NCC configuration and NMC management via a MIB. An RCST may be allowed to use several RCs. This requires the RC to also be configured with a set of TCs to classify the PDUs that arrive on the LAN Interface, associating each with one RC. The associated RC can use a NCC-assigned combination of capacity categories. To formulate a CR for the RC, the RCST measures a set of parameters to assess the capacity required. Examples include: the volume of queued traffic associated with the RC, the arrival rate at the RCST, and the transmission rate. This allows the RRM to provide a capacity requirement estimate in terms of additional/absolute volume and rate. An RC that supports several capacity strategies allows the RCST to make policy decisions about the ratio of capacity that it requests using each capacity category. For instance, both RBDC and VBDC methods may be used to request an allocation that corresponds to a certain number of timeslots per second, but may result in different allocation patterns, different costs, and different reactions to congestion. The NCC will usually associate resource limits with each RC (this may be configured via a MIB). The DVB-RCS2 Lower Layer Specification places limits on the requested level of resources. A simple instantiation may statically map one type of traffic to a specific RC and also associate this with a particular HLS Service. However, in general, there is not a need for a 1-to-1 mapping between use of RCs (for resource management) and an HL Service (for QoS management). Each CR message is identified by the RC (via the rc_index) when it is sent to the NCC. The NCC may in addition associate additional parameters with a specific RC (e.g. priority for allocation relative to other RCs, or timing requirements for allocated slots). In this way, time slot allocation process at the NCC may be tuned for a specific RC. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 81 There are several ways that RCs may be used in the RCST. Examples of use include: • In RCSTs that perform traffic classification (e.g. QoS based on DSCPs), an RC may realize a specific RRM scheme tailored to a traffic class. For instance, web interactive traffic may be requested using rate-based method (RBDC) while bulk and low priority traffic could use volume-based methods (VBDC). The allocation methods at the NCC may be designed to maximize utilization when several traffic classes are simultaneously used, and a combination of capacity categories may be more optimal than using one alone. • RCs can also be used to identify a dedicated resource for control signalling and could for example be used to protect assign a higher resource priority for signalling exchanges or to accelerate the control signalling exchange (e.g. decreasing time for session setup when an RCST is idle). • RCs can be used to virtualize the underlying service. They can provide hard segregation/isolation of flows assigned to different SLAs operating over the same RCST equipment. This is useful when a satellite interface is shared among several users or ISPs. Each ISP/user makes CRs using their own RC for the capacity they need, subject to their own SLA. Allocations by the NCC are identified so they may be associated with the corresponding RC. A refinement of this scheme allows unused capacity at an RCST to be made available in an allocation pool that allows other RCs to use this to maximize their performance. An NCC may associate a correspondence between an RC and allocation parameters set at the NCC. This allows an RCST to influence the allocation for specific flows. Examples of use include: • An RC may allow the allocations to be tuned to realize a maximum allocation delay. • An RC may indicate that predictive-allocation (beyond the rate requested in a CR) is useful and hence influence FCA allocation, allowing the NCC to preferentially allocate unassigned capacity to the RCST that use these RCs. • An RC may be used to specify a pattern of allocation within the frame (e.g. to spread allocations over time). For example, an NCC may support "feathering" of voice timeslots across a frame, instead of allocating all capacity in one large timeslot burst. This minimizes jitter by delivering voice packets smoothly and evenly in systems using a large Frame period (e.g. hundredths of ms), reducing the time traffic needs to be held in jitter buffers by VoIP gateways. • RCs may be used to implement a resource priority scheme. This protects capacity should the satellite resource become congested, ensuring that premium services receive allocations ahead of other services. This may also determine which services may be delayed or not allocated in the case of RRM congestion. Prioritizing flows only has benefit when the load is high, under normal conditions all RCs will receive allocation. Table 6.11 summarizes some examples of RC use, the way in which the RC is applied, and the benefits that may result. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 82 Table 6.11: Example uses of Request Classes RC usage type Applications Benefits DiffServ resource management Dynamic allocation IP QoS. Implicit classification of traffic based on PHB. Capacity allocated based on CRs that use capacity category tuned to traffic specification. Beyond dynamic allocation Associates extra params at NCC for the RC, e.g. to enable FCA (or set a CRA rate) for a minimum rate guarantee. Capacity allocated based on CRs, but NCC may provide more (e.g. a minimum guarantee for an RCST during low traffic.) Integrated Services resource management Dynamic RC configured to match IS request. NCC tunes RC properties based on signalling rather than CRs. Protection from congestion of satellite resource Protection for Premium Services. Protection for user categories. Premium services or groups of users are not degraded during resource congestion. Service Virtualization Hard resource partitioning SLA separation for different services at one RCST. Requires allocated timeslots to reference an RC. RCs permit creation of multiple SLAs on same physical link. Separation between capacity for each SLA. Reusing allocated capacity Link sharing among SLAs. Requires allocated timeslots to reference an RC. May share unused capacity, to increase user performance. Satellite Link management Signalling Capacity reservations for control flows. Separate management, Accounting. Management can be classified in a specific RC to reserve bandwidth for out-of-band signalling. Feathering (spreading burst allocation across the superframe) Associates extra params at NCC for the RC. Provides a smooth allocation (rather than bursty) for selected flows. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 6.8 Joint use of RA and DAMA access mechanisms | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 6.8.0 Introduction | This clause provides implementation guidelines material for the integration of Random Access Allocation Channels (RA-AC) and Dedicated Access Allocation Channels (DA-AC). The clause first presents a number of used for user traffic transmission on the return link via random access. Then, the rest of the clause presents two different means of integration: 1) Integration of DA-AC and RA-AC for user data transmission. 2) Integration of DA-AC and RA-AC for user data transmission and signalling transmission. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 6.8.1 RA use cases | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 6.8.1.1 RA cold start | RA Cold Start is the simplest scenario, when an RCST is logged onto a network, but is initially idle with no current (or infrequent) capacity allocations. This includes infrequent CRA control slots (e.g. following a period with no traffic from an RCST). If new traffic arrives while the RCST is in this state, the DA request-allocation cycle can introduce a significant delay before transmission. In addition, when traffic is still growing (e.g. Following a UDP DNS packet or TCP SYN), it is difficult to accurately predict the size of CR needed for the initial demand in terms of volume, rate and required time of allocation. This can result in DAMA slots introducing delay (when optimizing for allocation efficiency) or being allocated over-allocated (when optimizing for RCST performance). In Cold Start, the RA channel could be used to start transmission, until capacity becomes available using the DA channel. Traffic could then be switched from the RA to DA. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 83 |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 6.8.1.2 RA-DAMA top-up | The Top-Up use-case arises when there is a sudden (unpredicted) increase in traffic. This could occur due to a scene-change in video, or the start of additional traffic flows. This use-case would use the RA channel to temporarily provide extra capacity (top-up) until additional requested DA capacity is received or the traffic burst passes. This use-case may mitigate the impact of jitter and/or sudden changes in demand. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 6.8.1.3 RA-DAMA back-up | A variant of Top-Up arises when allocation varies (e.g. due to rain-fade, system load, or mobility movement, or loss of a CR). In this case an RCST could use a combination of RA and DA capacity. The usefulness of this case will depend on many factors including system dimensioning. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 6.8.1.4 RA IP queue | The impact of DAMA access delay depends on the type of applications using the service. The response time for short interactive applications is a key performance index with a recommended value of 0,5 seconds per page for interactive gaming and 2 seconds for web browsing [i.68] and [i.69]. Table 6.12: End user performance expectations Applications Target Response Time E-commerce, ATM 4 seconds Web Browsing 2 - 4 seconds Telemetry, Interactive Games 0,5 seconds QoS techniques allow a separate IP queue behaviour to be assigned for these applications or traffic types. Typical Internet QoS methods (e.g. the Differentiated & Integrated Services model) are designed to avoid significant re-ordering of packets within a flow. That is, they attempt to preserve per-flow order. These methods do not produce unusual pathologies when used with TCP. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 6.8.1.5 RA capacity requests | The RA channel could be used to send CRs more rapidly than relying on CRA or periodic control slots. This technique can be used singly or in conjunction with another method (This is not examined here). |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 6.8.1.6 RA for SCADA | There are at least two distinct types of SCADA that may be classified as Managed SCADA (scheduled polling for data) and Random SCADA (e.g. alarms and events triggered by data). One of the advantages of RA in this domain is its ability to accommodate the large number of terminals required in many SCADA systems. Unlike DA, the complexity of using RA does not increase rapidly with increasing number of terminals. Hence the network size may be only limited by the total offered traffic. RA is also suited for applications that require transmission of short packets. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 6.8.2 Integration of DA-AC and RA-AC for user data transmission | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 6.8.2.0 Overview | The diagram reported in Figure 6.23 provides a conceptual summary of the key steps for transmitting data over a return channel. Traffic arrives in the form of IP datagrams to a Traffic Management Policy module, which applies classification based on some criteria, such as QoS, packet size, resource availability, or the inferred congestion level over the RA-AC. Classified data are then forwarded to a scheduling block and enqueued in L3 buffers for transmission. The scheduler monitors the state of the queues, generating reports (periodically or event-driven) that can be used for requesting capacity from the NCC, as in the case of traffic to be sent over a DAMA channel. Independently, an encapsulator triggers the scheduler at intervals, e.g. once per each superframe, asking for data. Scheduling requests are generated based on the amount of resources available for transmission at the terminal, and result in the scheduler forwarding a proper number of SDUs to the Encapsulator, which in turn fragments, encapsulates and passes them to the modem. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 84 Figure 6.23: Generic traffic transmission scheme for the return link at a terminal If data packets can only be sent over the DA-AC, as in 1st generation DVB-RCS systems, the described scheme is instantiated by having the traffic management module simply classify traffic, e.g. into EF, AF and BE priorities, and by having data be enqueued in possibly separate buffers. The state of the buffers drives future capacity requests, while the encapsulator draws datagrams based on currently available DAMA resources. Should it not be possible to send an IP datagram completely within the available capacity, the spare fragments are processed in subsequent superframes. On the other hand, ETSI EN 301 545-2 [i.1] foresees the possibility to map PPDUs to different ACs, enabling the additional degree of freedom of splitting traffic between DA-AC and RA-AC. In this perspective, traffic sent through random access procedures is likely to experience lower delay by avoiding the capacity request process of DAMA, while experiencing good success rates by virtue of advanced MAC schemes such as CRDSA. Conversely, the terminal access rate limitation and the load control algorithm provide maximum boundaries for RA channels that suggest their usage for short bursts of data. These remarks hint a first possible approach for the integration of RA- and DA-ACs, relying on the former to send small or urgent datagrams as well as signalling, while reserving the latter for more predictable and less delay-sensitive traffic. Such a solution can also be flanked by a more dynamic strategy, as data could be split among channels taking into account current conditions in terms of load and congestion. In such an approach, RA data could also be opportunistically reallocated to DA slots if exclusive resources exceeding the current amount of DAMA datagrams are available. A high level description of the discussed integration strategies is provided in Figure 6.24. The remainder of this clause, instead, describes in greater details different solutions that stem from these core ideas, specifying the required modifications to the basic scheme of Figure 6.23 as well as discussing possible advantages and drawbacks. Figure 6.24: Conceptual approach for RA- and DA-AC integration |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 6.8.2.1 Design requirements | An algorithm for the integration of RA and DA-AC should fulfil the following requirements: 1) The algorithm should take higher layer requirements into account (e.g. QoS and application requirements, such as avoiding out-of-order delivery at transport layer or large variations in delay variation). 2) The algorithm should take lower layer requirements into account (e.g. the size of L3 packets to limit the number of L2 fragments and the total L3 packet error rate). 3) The algorithm should take the current load of the DA channel and the RA channel into consideration. 4) The algorithm should take the access rate limitation mechanism on the RA channel into consideration. 5) The algorithm should allow L3 packets, which were originally intended for the RA-AC to be opportunistically relocated to the DA-AC, in case DA capacity is still available after serving all DAMA queues (i.e. DA-AC excess capacity). ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 85 6) The algorithm should not allow L3 packets that were originally intended for the DA-AC being opportunistically relocated to the RA-AC, because this would reduce the accuracy and consistency of the DAMA rate and queue measurements, which are used for the capacity request generation. Furthermore, it would counteract the first requirement since such a relocation could cause unacceptably delay variations or unacceptably deep misordering of L4 segments. 7) The algorithm should be able to improve the overall delay while satisfying the aforementioned requirements. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 6.8.2.2 Example algorithm | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 6.8.2.2.0 Algorithm Description | Taking into account the features specified in clause 6.8.1.1, in this clause a possible algorithm for the integration of RA and DA-Acs is proposed. As discussed, higher layer requirements should be considered. In the first place, some higher layer applications may result sensitive to out of order delivery of L3 IP datagrams and lose in performance (e.g. TCP). Other issues may affect applications which are sensitive to high delay jitter (e.g. issues with the playback buffer of VoIP). Problems of out of order delivery could typically result if data is sent jointly over DA-AC and RA-AC with opportunistic relocation in both directions. Problems of higher delay jitter could typically result if data is sent over the RA-AC or jointly over DA-AC and RA-AC with opportunistic relocation in both directions. In particular, for the joint use of DA-AC and RA-AC a higher variation of delay jitter might result due to the different delays of DA and RA and data might arrive not in order either due to the different DA and RA delays (queuing and channel access delays) or due to the order of decoding RA and DA sections of a superframe. For this reason, a first classification is applied to L3 PDUs to decide which packets should go over the DA-AC (i.e. the packets affected by the aforementioned issues) and which packets are eligible to go either over DA-AC or over RA-AC. Since this classification is based only on requirements coming from the higher layers, it is denoted as Higher-Layer- Classification (HLC). Figure 6.25 shows this block at the left and also all the other blocks discussed in the following. Figure 6.25: Proposed RA-DAMA integration strategy Among all the packets that are tagged as eligible to be sent over the RA-AC or the DA-AC, a second classification is applied, which takes limitations coming from the lower layers (Lower Layer Classification, LLC) into account. Here, the limitation is basically the size of the L3 PDU. The justification for this is that large L3 PDUs will result in a large number of L2 PPDUs in the fragmentation process. If such a large number of L2 PPDUs is sent over the RA-AC (which naturally has a higher loss rate than the DA-AC) the probability of losing the L3 PDU increases. For this reason, a threshold is applied, so that packets above a given size are tagged to be sent only over the DA-AC, whereas the others remain eligible for RA-AC transmission. The value of the classification threshold depends on the tolerable packet loss rate as well as on the employed contention scheme on the RA-AC, i.e. Slotted Aloha (SA) or CRDSA. Once the set of packets that may be sent over RA-AC satisfying HLC and LLC has been determined, a final choice is made to identify which of them should actually be enqueued in RA buffers. This decision aims at reducing the experienced delay. Therefore, the algorithm estimates the expected delay that a packet would undergo if sent over the DA-AC (i.e. taking into consideration the current DA load and queue state) and compares it to an estimation of the expected delay that the same packet would undergo if sent over the RA-AC (taking into consideration the current RA load and queue state as well as and restrictions given by the load control and the terminal access rate limitation settings). ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 86 If the expected delay over the RA-AC is lower than the one over the DA-AC, the packet is enqueued in the RA queue. If the expected delay for the RA-AC is equal or bigger to the DA-AC, then the packet is assigned to the DA-AC. The RLE encapsulator receives information about the DAMA capacity that the terminal was assigned in the current superframe. As long as DAMA capacity is available, it triggers the scheduler to send further L3 DAMA packets. Should the DAMA queues be empty but still capacity available, then an opportunistic relocation of packets from the RA queues towards the DAMA queues goes into effect. This is done to utilize the DAMA capacity assignments and lower the load and traffic over the RA-AC. The methods for estimation of the RA and DAMA delay are explained in more detail clause 6.8.2.2.1. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 6.8.2.2.1 DA-AC delay estimation | The only information which a terminal has available to estimate what will happen in the future is represented by the capacity requests it sends to the gateway. With this and the knowledge of the propagation delays, the terminal can estimate when the assignment should arrive in the future. The terminal furthermore has to assume that it will receive the capacity that it has requested, since it has no information about the overall system load status, which could indicate that a request may not be served fully due to overload. On the other hand, assuming that the requested capacity will be also assigned later on holds in low to average network loading conditions and may result in an overestimation of the serving rate only in situations where the network is getting saturated and overloaded. Such an overload should happen only seldom and not in nominal conditions if the network is properly dimensioned. The approach for the DA-AC delay estimation is then the following: a) Track all stated capacity requests within the terminal and keep book of them. b) Compute for every future SF the capacity that the terminal expects to get assigned from the table of capacity requests it has sent earlier and considering the propagation delay until the assignment is expected. c) Compute the number of bytes waiting in the queue ( DA L ) in front of the current L3 PDU ( 3 L PDU L ). d) Integrate the expected capacity assignments from b) until the total number of bytes ( DA L + 3 L PDU L ) has been reached. The end time of the superframe in which the integration exceeds this value is then the expected delay (excluding the propagation delay which is a constant adding on top) of the current L3 PDU. Table 6.13 shows an example of a possible Capacity Request storage table in a terminal. Table 6.13: Example for capacity request storage table in the terminal Type Start End Capacity CRA 1T ∞ 1 C (Bytes/SF) RBDC1 2T 3T 2 C (Bytes/SF) RBDC2 4T 5T 3 C (Bytes/SF) VBDC 3T 3T 4 C (Bytes/SF) VBDC 6T 6T 4 C (Bytes/SF) In Table 6.13, a Constant Rate Assignment (CRA) starting at 1T is noted. As is the nature of CRA, the assignment is constant without a defined end. In the case of an update the CRA value in Table 6.13 needs to be updated as well. If CRAs are expressed in a byte rate per second (instead of number of bytes per SF), then a normalization to the SF duration needs to be done: 1 rate SF C CRA T = ⋅ ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 87 In Table 6.13, the two RBDCs are sent at different times. The first one requests 2 C Bytes per SF (if expressed in as byterate, a normalization to the SF duration as shown for the CRA needs to be done) in the time period starting at 2T and lasting until 3T . The second RBDC requests an assignment of 3 C Bytes/SF in the time interval 4T until 5T . Finally, two VBDCs are sent to get granted in the SFs starting at time 3T and 6T , both asking for a volume of 4 C Bytes. Figure 6.26 illustrates these requests over time and shows the resulting expected capacity assignment total C for every superframe, provided that all capacity requests are fully granted by the NCC without delay. 1 C 2 C 3 C 4 C total C Figure 6.26: Example of CapReqs vs. time ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 88 Once the integral over total C in Figure 6.26 exceeds the number of bytes stored in the queue + the current L3 PDU size, the SF in which the current L3 PDU is expected to be transmitted is found, and the expected queuing delay can be computed with the number of required SFs and the superframe duration. DA SF SF N T Δ = ⋅ |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 6.8.2.2.2 RA delay estimation | The serving rate is estimated from the signalled load control and access rate limitation values directly. Let: • Nmax = maximum number of fragments that can be sent per RA block. • Lmax = maximum number of successive RA blocks that can be accessed. • pb = backoff probability. • I = number of idle RA blocks that the node should not access when rate access limitations have been reached. • T_block = the duration of a RA block. Ignoring backoff limitations, the maximum average rate at which fragments (i.e. unique payloads) can be transmitted can be estimated as: max max max ( ) block N L L I T μ = + From this, the average number of fragments per transmission opportunity that can be sent considering backoff, i.e. the predicted service byterate, can be computed as: R^ = (1-pb)µ This value of R^ is used to predict the delay undergone by a datagram if enqueued in a RA-AC buffer as: ^ 1 ^ ^ ) ( b i n i b Q R F R F t T p = + = , where: np = number of datagrams currently enqueued in buffer Q. F = number of fragments of which the datagram to be enqueued is composed. Fi = number of fragments of which datagram i is composed. 6.8.3 Integration of DA-AC and RA-AC for user data and signalling transmission The RA-DAMA integration strategy described in the previous clause focuses on integrating the transmission of user data, while the signalling (e.g. capacity requests) is sent either in unsolicited dedicated control slots or in state-of-the art Slotted ALOHA (SA) channels. The provision of a RA-AC for data transmission may have a positive effect on the overall delay. In this clause, another interesting application case for a RA-AC is investigated, namely to replace the SA minislot signalling carrier with a more efficient CRDSA RA-AC which may also be utilized for user data transmissions. Doing this brings two main benefits: 1) The spectral efficiency of the signalling carrier increases compared to SA, allowing more signalling to be carried in the same amount of bandwidth. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 89 2) The CRDSA signalling carrier can be also used for additional user data transmission in addition to the signalling which may further increase the efficiency. In the following, the focus is on the case of having signalling sent over a RA-AC (referred to as RA-SIG-AC) together with additional data traffic. From a practical implementation point of view, the fact that the signalling messages are L2 packets while data are L3 packets needs to be taken into account. This means that the L2 Signalling and the L3 IP packets cannot be enqueued in the same buffer but need separate ones. Figure 6.27 illustrates the integration of the L3 IP data with the L2 signalling data over a common RA-SIG AC. Figure 6.27: Integration concept of IP data and L2 signalling over a common RA-SIG allocation channel Since the signalling messages are L2 messages, there is no L3 queue existing, but signalling messages are directly enqueued in a L2 queue at the encapsulator upon their arrival. The L3 IP data on the other hand are enqueued in a L3 buffer. The encapsulator triggers the L3 scheduling of a new L3 packet whenever there is remaining space in the RA-SIG-AC. Once the scheduled L3 packet arrives, its fragments are enqueued in a separate L2 buffer. This is done since it is considered meaningful to give signalling messages priority over user data packets, since blocking signalling messages will result in delayed or missing DAMA allocations, which is not desirable and has a negative impact on the DAMA performance. Keeping separate queues and scheduling them with a priority scheduler is foreseen to ensure the priority of scheduling over data messages. It should be noted that only the L2 fragments of one L3 packet may reside in the L2-DATA-BUF queue, while in the L2-SIG-BUF queue several different signalling messages may be stored. The L2 PPDU which is selected for transmission by the L2 scheduler is then stored in the transmission buffer. There is only a single transmission buffer for the RA-SIG-AC. A new L2 scheduling round only starts if the current transmission buffer is empty. Figure 6.28 illustrates the procedure to be followed whenever a new L2 signalling message arrives. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 90 Figure 6.28: Flowchart for actions upon arrival of a L2 signalling message As explained before, the L2 signalling message is in any case enqueued into the L2-SIG-BUF. In case the Transmission Buffer (TxBuf) is empty at this time, a new L2 scheduling round is initiated immediately. If there is still data pending in the TxBuf, then the procedure finishes. The procedure upon arrival of a L3 data packet, scheduled by the L3 scheduler is the same (see Figure 6.29), except that the IP data are enqueued in a separate buffer called L2-DATA-BUF in Figure 6.27. Figure 6.29: Flowchart for actions upon arrival of a L2 signalling message Transmission over the RA-SIG-AC can be triggered by two events: • A SF activation event: Occurs at the start of a new SF. • A L2 scheduling event: Occurs whenever the L2 scheduler is triggered. Figure 6.30 illustrates the flow chart upon a SF-activation event. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 91 Figure 6.30: Flow chart upon a SF activation event Whenever the time of activation of a new SF has arrived, this procedure is executed. First the status of the TxBuf is checked. In case the TxBuf is currently empty, a new round of the L2 scheduler is triggered and the procedure is finished: • If there is data in the transmission buffer, it needs to be checked whether the fragments in the buffer are part of an already initiated transmission of an existing higher-layer PDU (i.e. pending data) or whether they are the fragments of a newly arrived higher-layer PDU. This distinction is necessary to comply with the RA load control specification for the following reason: Set the case that the L2 scheduler delivered a new L2 message in the TxBuf. According to clause 9.7.3.2 in ETSI EN 301 545-2 [i.1], at this time a decision about the backoff needs to be done. According to ETSI EN 301 545-2 [i.1], the decision can only be made if the data is already received, i.e. located in the TxBuf, not before. In case a backoff was decided, then the SF-activation procedure needs to check the expiry of the backoff time. • Set another case where data is in the TxBuf, but this data is part of a transmission initiated already earlier but which could not yet be finished. In this case, the expiration of the backoff counter is not relevant, but a decision about avoiding the usage of the current transmission probability needs to be made. In order to differentiate between the two cases, a flag for indication of pending data (Pending Data flag) is introduced. Other ways of implementing the distinction are possible as long as the procedure complies with the load control specification. In case the data payload currently in the TxBuf is pending data of an earlier transmission, the sendPendingTxBuffer subprocedure then checks whether the RA transmission opportunity needs to be omitted or not, i.e. if the maximum number of consecutive blocks is not yet exceeded, the maximum number of unique payloads per block is not yet exceeded and the random decision for avoiding the transmission opportunity was in favour of transmission, then the payload is sent. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 92 In case the full payload could be sent, the TxBuf will be empty. In this case another L2 scheduling round is triggered to exploit the full capacity of the RA-AC. If data is still in the TxBuf then the load control limitations have been reached and the current assignment round is over. In case the data in the TxBuf is a new payload, then the procedure checks whether the backoff time has expired or not. In case it has not yet expired, the BackoffTimer is decremented and the current allocation round is finished. In case the backoff has expired, the new payload is sent over the RA-AC (insertRaDa block), where again the check of the load control limitations as for the sendPendingTxBuffer is applied. In case the payload could be sent completely, then also here another L2 scheduling is initiated. If the payload could not be sent completely, the pending data flag is set and the current allocation round is over. The procedure of a new L2 scheduling round is depicted in Figure 6.31. Figure 6.31: Flowchart of the procedure upon a new L2 scheduling round For the L2 scheduler, different scheduling strategies can be implemented. Here a strict priority scheduler is proposed. This scheduler always assigns priority to signalling packets for the aforementioned reasons. In other words, whenever data is waiting in the L2-Sig-Buf and the L2-Data-Buf, the signalling messages will always be scheduled first. After scheduling a message (signalling or data) and enqueuing in the TxBuf the decision of the backoff is made. In case the transmission should be backoffed (with probability pb), the backoff time is set to the back_off_time specified in the received signalling fields and the scheduling round is over. In case of no backoff, the insertRaDa function checks whether the load control limitations (max number of consecutive payloads, max number of unique payloads per block) are exceeded and in case transmits the payload over the RA-AC. If some data is still remaining in the buffer, the pending data flag is set and the scheduling round is over. If everything could be sent, another L2 scheduling round is initiated (recursively) and the current allocation round is over. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 93 |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 7 M&C Functions Supported by L2S | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 7.1 Control of EIRP | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 7.1.0 General | The manufacturer of the RCST needs to state precisely the operating range of the EIRP control of the RCST. This is because different system operators may use different strategies for RCST EIRP control. In some system designs, a wide rain fade margin may be expected and tight RCST uplink power control exercised. In other system designs, RCST uplink power control may not even be used depending on system cost trade-offs. RCST EIRP control may be exercised by the RCST itself or by the NCC. It is generally anticipated that in most system designs the RCST EIRP will be adjusted up or down in nominal 0,5 dB increments over the operating EIRP range of the RCST by commands from the NCC. This will be in response to direct or indirect measurements of the link margin of the RCST in question. For large step changes in EIRP level it is unreasonable to expect the RCST to provide a nominal 0,5 dB accuracy and so in this case the specification only calls for the resulting power change to be within 20 % of the dB value of the requested step change. In this circumstance it is expected that the EIRP step change will be followed by incremental up or down EIRP changes in nominal 0,5 dB steps. Whenever the EIRP of the terminal is increased, there is a possibility for spectral regrowth (for linear modulation) to occur, impacting the performance of the adjacent channels. There is a number of different approaches to solve this problem. These are left to the system designer. One possibility is that the OBO could be controlled in conjunction with the NCC. The NCC determines from time to time the operating point on the output power versus input power curve of the HPA. This can be realized for example by requesting RCST Transmit Power headroom as described below. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 7.1.1 RCST Transmit Power Headroom | This clause provides some considerations of the "power headroom" parameter, including why it is useful and how the value can be determined. The "headroom" parameter reported by the RCST in the control PPDU (see Table 8-8 of ETSI EN 301 545-2 [i.1]) indicates the difference (in dB) between the actual transmit power and the maximum possible. The main rationale for reporting this parameter is to aid the decision process in systems that employ countermeasure techniques against fading or other channel variations. The most reliable signal quality indicator at the NCC is typically the C/N of the received bursts. However, if the back-off of the RCST is not known, the NCC cannot easily determine the RCST's current capability. Schemes that attempt to track the back-off by accumulating the commanded power variation can easily get out of step, if messages are missed or if saturation occurs in places other than expected. If the headroom H is known explicitly, the capability of the RCST can be expressed as: ( ) H R N C N C S + + = 10 max 0 log 10 The RCST will be able to use any return link transmission mode that has a threshold C/N0 not exceeding C/N0|max. This is particularly an issue at times where the channel conditions are improving, such as at the tail end of a fade event: If the headroom is not accounted for in these situations, there is a risk that the RCST will continue simply to back-off the power. If this happens, the C/N never improves and the RCST may thus never switch to a higher-rate and/or more efficient mode of transmission. This is generally less of an issue while conditions are worsening; power control will tend to maximize the power so the headroom is usually small at that end of the process. These considerations are what led to inclusion of headroom reporting in the RCS2 return link signalling. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 94 The headroom can be determined at the RCST in a number of different ways, with different accuracy and cost implications. In the simplest form, the headroom can be estimated from the IF attenuator setting. This requires no additional hardware; in particular, it does not require any communication between the modem part and the power amplifier proper. However, this method does not allow for the compression in the power amplifier, so it will tend to be less accurate when operating close to maximum power. The method can be improved for example by using a look-up table that models the compression characteristic. However, in order to bring a meaningful improvement, such a table really needs to be calibrated for individual power amplifiers - the production spread of characteristics can be quite considerable. This complicates the installation, or at least forces the pairing of modem (indoor unit, IDU) and power amplifier to be made at the factory so they can be calibrated together. This may be acceptable in many situations, but gives complications later if either of the units needs to be replaced. In any case, the look-up table method does not account for changes due to temperature variations and aging of the power device, which will alter the relationship between the IF level and the position of the compression curve. The most reliable method of determining the headroom is obviously to have a power detector as part of the RF amplifier unit. The power detector itself can be a very simple and cheap circuit; complications and cost arise mainly from the need to communicate the measurements to the modem/IDU. This is particularly the case for stand-alone power amplifiers (BUC's). These typically have a single-cable connection that carriers DC power, IF signal and a reference tone for the up-converter. Adding the capability to signal back from the BUC to the IDU, for example using DiseqC, necessitates bulky and expensive diplexers as well as other additional hardware. This is the main reason explicit power detection is not commonly used in low-cost units. It is easier to accommodate the power measurement signalling in more integrated RF units, for example those that combine transmit and receive functions ("transceivers"). It is typically much easier and cheaper to accommodate the signalling on the receive cable, which does not carry high levels of power. Such more sophisticated units may already have a signalling path for other monitoring functions. Fully-integrated terminals, in which the modem and RF devices are in the same enclosure, of course have no such issues with reporting of the measured power. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 7.2 Pointing Alignment Support | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 7.2.1 Point Alignment Support descriptor | The descriptor contains a single data structure the contents of which depends on its type, which is indicated in the alignment_control_type field. The alignment_control_type field values are listed in Table 7.1. Depending on this field value, the descriptor may be encapsulated in TIM-B or TIM-U signalling tables. The NCC may encapsulate multiple pointing alignment support descriptors in the same TIM-U or TIM-B signalling tables. Table 7.1: Alignment control types ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 95 With alignment control type of "0", the NCC may broadcast whether or not pointing alignment is required in the network, whether or not the NCC supports CW-based return link alignment, whether or not the NCC supports IB-based return link alignment, and the forward link SNR threshold that should be exceeded to complete forward link alignment procedure. With this alignment control type, the NCC may also convey information pertinent to a user-defined alignment procedure. With alignment control type of "1", the NCC may broadcast an ASCII string that provides a reference (e.g. a phone number) to a Network Operations Centre, which may provide assistance with the return link alignment procedure. With alignment control type of "65", the NCC may unicast to a RCST the remaining duration in which the RCST may continue with return link alignment procedure and the threshold values for co-polar and cross-polar measurements. Upon reception of this descriptor, the RCST resets its remaining duration timer to the number of seconds indicated in the remaining duration field of this descriptor. With alignment control type of "64", the NCC may unicast to a RCST an alignment population ID value that is different from the RCST's operational population ID in addition to the information that can be conveyed with the alignment control type of "65". Upon receiving this descriptor, the RCST tunes to the forward link signalling service identified by the alignment population ID in the RMT table. With alignment control type of "66", the NCC may unicast to a RCST a 16-bit pattern that the RCST transmits in installation bursts during the IB-based return link alignment. DVB-RCS2 dictates using the Logon SDUs in dedicated access logon slots allocated by the NCC to transmit installation bursts. RCST replicates the 16-bit pattern as many times as necessary to fill in the payload of the Logon SDU for installation burst transmission. Clause 8.3.1 of ETSI EN 301 545-2 [i.1] elaborates further on the use of Logon SDUs for installation burst transmission. The logon slots dedicated for installation burst transmission are allocated by the NCC in the BTP. With alignment control type of "68", the NCC may unicast to a RCST the start time, the duration, and the frequency values that the RCST should use during the CW-based return link alignment procedure. With alignment control type of "67", the NCC may unicast to a RCST the EIRP that the RCST should use during the CW-based return link alignment procedure in addition the information that can be conveyed with the alignment control type of "68". Note that devices that cannot dynamically control their EIRP should terminate the return link alignment procedure with failure upon receiving this descriptor. The RCST may provide visual aids to the human operator indicating the reason for failure. With alignment control type of "96", the NCC may unicast to a RCST the CNR, co-polar, and cross-polar measurements read on the return link transmission from the RCST. In addition, the NCC may convey alignment status of the RCST, which may be "in-progress", "failure", and "success". The RCST should continue the return link alignment procedure for as long as its alignment status is "in-progress" provided that the most recent remaining duration value from the NCC has not been exceeded. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 7.2.2 Elements in Logon SDU and Logon Response Descriptor | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 7.2.2.1 Logon SDU | The Logon SDU contains a 4-bit entry_type and a 4-bit access_status fields. The RCST may request for alignment support by assigning 0x00 to the entry_type field in the Logon SDU. When the Logon SDU is used to transmit an installation burst during the return link alignment, the RCST assigns 0x05 to the entry_type. If the least significant bit in the access_status field is '1', the RCST indicates that the NCC has confirmed earlier that the RCST has completed alignment procedure. The NCC may disregard this indication and enforce the RCST to repeat the alignment procedure. When the Logon SDU is used to transmit an installation burst during the return link alignment, the RCST assigns 0x00 to the access_status. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 96 The Logon SDU may contain a number of Logon Elements after the entry-type and access_status fields. Each Logon Element may be of variable size. Each Logon Element contains a 4-bit type and a 4-bit size fields. Logon element of type '10' contains RCST pointing alignment support capabilities. This is a 3-byte logon element that indicates the nominal EIRP in the pointing direction if the RCST is a fixed-EIRP device. The logon element may also be used to indicate that the RCST supports CW-based, IB-based, and dynamic-EIRP return link alignment procedures. When the Logon SDU is used to transmit an installation burst during the return link alignment, it only contains logon elements of type '11'. In this case, the Logon SDU encapsulates as many logon elements as necessary to fill in the space in the logon slot. If the remaining space is only 1 byte, then the RCST fills in this space with the most significant byte of the 16-bit pattern indicated in the Pointing Alignment Support descriptor from the NCC. Figure 7.1 summarizes the Logon SDU with highlights over the fields regarding the pointing alignment support. Figure 7.1: Logon SDU elements for pointing alignment support |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 7.2.2.2 Logon Response descriptor | The Logon Response descriptor contains a 4-bit RCST_access_status field. This field has the same encoding with the access_status field in the Logon SDU. If the least significant bit of the RCST_access_status field is '1', then the NCC confirms in the Logon Response descriptor that the RCST does not need to repeat the pointing alignment procedure. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 7.2.3 Example signalling exchanges | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 7.2.3.1 Continuous-wave (CW) alignment | Figure 7.2 shows an example CW-based return link alignment message exchange sequence between the RCST and the NCC. It is assumed that the RCST is a fixed-EIRP device, which should be indicated in the Logon SDU in the pointing alignment support indicator logon element. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 97 Upon power-up, the RCST tunes to the start-up TDM to download periodical NIT, RMT, and TIM-B messages. The TIM-B messages from the NCC indicate that the NCC requires RCST pointing alignment support before joining in the network, that the NCC supports CW-based return link alignment, and the minimum forward link SNR threshold that should be exceeded before the return link alignment procedure may commence. The RCST installer adjusts the antenna alignment during forward link alignment before initiating the return link alignment procedure. Upon completing the forward link alignment, the RCST finds and tunes to its FLS TDM stream corresponding to its population ID. Upon reception of SCT, FCT2, BCT, and TBTP2 tables, the RCST sends a Logon SDU in a RA Logon slot. The Logon SDU indicates that the RCST requests pointing alignment support and that the NCC has not confirmed an earlier alignment completion. In response, the NCC sends a TIM-U containing two descriptors. The first is a Logon Response descriptor with RCST_access_status==0x00. It is recommended that the NCC does not assign any RCS-MAC address to the RCST at this stage, because it is not certain yet that the return link alignment procedure will succeed. Note that lower layer unicast signalling on the forward link always uses the 48-bit RCST hardware identifier as the unicast address. The second descriptor in the TIM-U is a Pointing Alignment descriptor containing alignment_control_type of '66' indicating that the NCC supports CW-based return link alignment procedure. This is accompanied with the frequency, start time, and duration for the CW signal that the RCST needs to transmit. The NCC may send another Pointing Alignment descriptor before the CW transmission is over, effectively, extending the duration of the CW transmission (this is not the case in the example). In response to the CW transmission, the NCC sends a TIM-U with two Pointing Alignment descriptors. One of the descriptors provides feedback in terms of CNR, co-polar, and cross-polar readings. The second descriptor instructs additional CW transmission. In response to the second CW transmission from the RCST, the NCC sends a TIM-U containing a Pointing Alignment descriptor that indicates return link alignment success. This is followed by the RCST accessing the RA channel to send a logon SDU. In response, the NCC encapsulates in a TIM-U a Logon Response descriptor and a Control Assign descriptor. The Logon Response descriptor assigns all necessary addresses to the RCST since the return link alignment has been achieved. The Control Assign descriptor allocates to the RCST a number of periodic Control SDU slots for achieving and maintaining fine synchronization in addition to other return link M&C signalling. The TIM-U may at this stage also contain descriptors necessary to fully configure the RCST. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 98 Figure 7.2: Example message exchange with CW-based return link alignment, fixed-EIRP RCST, the same population ID for operational and alignment phases |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 7.2.3.2 Installation-burst (IB) alignment | Figure 7.3 shows an example IB-based return link alignment message exchange sequence between the RCST and the NCC. It is assumed that a different alignment population ID is maintained than the operational population ID. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 99 Upon power-up, the RCST tunes to the start-up TDM to download periodical NIT, RMT, and TIM-B messages. The TIM-B messages from the NCC indicate that the NCC requires RCST pointing alignment support before joining in the network, that the NCC supports IB-based return link alignment, and the minimum forward link SNR threshold that should be exceeded before the return link alignment procedure may commence. The RCST installer adjusts the antenna alignment during forward link alignment before initiating the return link alignment procedure. Upon completing the forward link alignment, the RCST finds and tunes to its FLS TDM stream corresponding to its population ID. Upon reception of SCT, FCT2, BCT, and TBTP2 tables, the RCST sends a Logon SDU in a RA Logon slot. The Logon SDU indicates that the RCST requests pointing alignment support and that the NCC has not confirmed an earlier alignment completion. In response, the NCC sends a TIM-U containing three descriptors. The first is a Logon Response descriptor with RCST_access_status==0x00. It is recommended that the NCC does not assign any RCS-MAC address to the RCST at this stage, because it is not certain yet that the return link alignment procedure will succeed. Note that lower layer unicast signalling on the forward link always uses the 48-bit RCST hardware identifier as the unicast address. The second descriptor in the TIM-U is a Pointing Alignment descriptor containing alignment_control_type of '64' indicating the use of an alignment population ID that is different from the RCST's operational population ID. The third descriptor is another Pointing Alignment Support descriptor with alignment_control type of '66' indicating the 16-bit pattern to be used in installation bursts. The RCST searches in the RMT the FLS TDM that corresponds to the alignment population ID and tunes to this TDM and acquires the FLS. The BTP (SCT/FCT2/BCT/TBTP2) in this FLS contains DA logon slot allocations for the RCST. The RCST sends a Logon SDU with alignment probe (installation burst) in the DA logon slots allocated by the NCC. In response, the NCC sends a TIM-U with a Pointing Alignment descriptor providing feedback in terms of CNR, co-polar, and cross-polar readings. TIM-U also contains a Correction Message descriptor exploiting the return link alignment procedure also for fine synchronization achievement purposes. In parallel, the RCST receives additional DA logon slot allocation in TBTP2. The RCST sends additional Logon SDU in the DA logon slots. In response, the NCC sends a TIM-U containing a Pointing Alignment descriptor that indicates return link alignment success. In addition, the TIM-U contains a Correction Message descriptor and a Control Assign descriptor. In this specific example, it is assumed that the fine synchronization is not achieved yet, even though the return link alignment procedure is successful. The RCST uses the Control SDU slots allocated by the Control Assign descriptor for fine synchronization. When the fine synchronization is achieved, the NCC sends a TIM-U with Satellite Forward Link and Satellite Return Link descriptors in addition to a Correction Message descriptor. The Satellite Forward Link and Satellite Return Link descriptors correspond to the operational population ID of the RCST. The RCST tunes to the operation TDM FLS and seeks for RA logon slots in the BTP (SCT/FCT2/BCT/TBTP2). It is possible that the NCC may allocate DA logon slots for the RCST to speed up the re-logon process after alignment. The RCST sends a Logon SDU with entry_type==0x1 and the access_status having '1' as the least significant bit. In response, the NCC sends a TIM-U containing a Logon Response descriptor and a Control Assign descriptor. The Logon Response descriptor assigns all necessary addresses to the RCST since the return link alignment has been achieved. The Control Assign descriptor allocates to the RCST a number of periodic Control SDU slots for maintaining fine synchronization in addition to other return link M&C signalling. The TIM-U may at this stage also contain descriptors necessary to fully configure the RCST. Note that the NCC may allocate a sequence of DA logon slots in the BTP. If necessary, a much longer superframe duration may be used to allocate longer DA logon slots. Also note that the NCC may terminate the alignment procedure by sending a "failure" in the Pointing Alignment Support descriptor. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 100 Figure 7.3: Example message exchange with IB-based return link alignment, a different population ID for operational and alignment phases ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 101 |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8 Transmission Security Implementations | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.1 Introduction | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.1.0 General | This clause describes three TRANSmission SECurity (TRANSEC) implementations for the DVB-RCS2 waveform, each covering a different market or profile emphasis. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.1.1 Scope of TRANSEC Protection for DVB-RCS2 | In the context of DVB-RCS2, TRANSEC protection is defined to include: • Protection of Channel Activity Information - To disguise user traffic volumes in both forward and return channels. • Protection of Control and Management Information - To encrypt or otherwise disguise control plane and management plane signalling and signalling tables in both forward and reverse channels • NCC and RCST Authentication - To ensure that only valid RCST can log-on to a valid NCC and conversely that neither an invalid RCST cannot log on to a valid NCC, or a valid RCST cannot log-on to an invalid NCC. • Anti-Jam and Low Probability of Intercept - These attributes are considered to be at a lower priority with respect to the previous points and are omitted for the present. The countermeasure techniques commonly used for mitigating the above risks consist of link-layer encryption, authentication and traffic activity concealment / obfuscation (and associated key management). Of all these, the link layer encryption is foremost. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.1.2 TRANSEC Profiles | The implementations to be described here are based on the following DVB-RCS2 TRANSEC profiles: • Consumer • Professional • Governmental As the naming suggests, each profile has a different market emphasis in mind, and accordingly supports a different balance of TRANSEC features. However, features from each implementation here (and others) can be combined to create Custom profiles. The TRANSEC profiles discussed here have several common features such as the use of AES-256 algorithm for link layer encryption, but differ primarily in cryptographic modes of operation and approaches to key management and authentication. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 102 |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.1.3 Generic Implementation Approach | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.1.3.1 Security Architecture | The DVB-RCS2 waveform introduces the notion of TRANSEC hooks. The "hooks" allow the extension of normative DVB-RCS2 waveform with TRANSEC countermeasures. The hooks include: • What and where to encrypt. • What to authenticate. • Where to introduce dummy traffic. • New signalling and reserved fields in control messages for TRANSEC management. The "hooks" introduced to the air interface design should accommodate different security implementations with minimal or no impact to complex intellectual property blocks used for modulation/coding and demodulation/decoding and the DVB-RCS2 HLS [i.4]. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.1.3.2 Link Layer Encryption Hooks | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.1.3.2.0 General | TRANSEC encryption should be applied to both Forward Link (FL) and Return Link (RL) of a DVB-RCS2 network, to all data packets (payloads and headers) and to all signalling, with the following exceptions: • The predefined synchronization sequences, when used (e.g. burst preambles), should be sent in the clear, in order to allow the synchronization of the demodulators. • The MODCOD field in the Physical Layer (PL) header of the DVB-S2 frames should be transmitted in the clear in order to facilitate frame demodulation / decoding. • The CRC trailers in DVB-S2 frames should be sent in the clear. Link layer encryption should be based on approved algorithms, e.g. AES-256, used in approved modes of operations. For governmental applications the encryption algorithms and the operation modes should be government approved. As a minimum, the Forward and Return Link should each use one Encryption Key for the Interactive Network. It is also acceptable for individual Remote Terminals, or groups of Remote Terminals, to each possess one or more Encryption Keys for each link. Continuous-carrier return and mesh links should use the methods defined for forward link transmission. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.1.3.2.1 Forward Link Encryption | Regardless of the precise technique used for Forward Link encryption, each DVB-S2 frame should carry an initialization Vector (or Initial Counter Value) that can be used for decryption of the frame. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.1.3.2.2 Return Link Encryption | In the interest of bandwidth efficiency, Return Link and Mesh Link TDMA transmissions do not need to carry explicit initialization vectors. Instead, unique initialization vectors for these transmissions can be computed from a combination of shared secrets and (extended) identifiers of the superframe, frame and time slot already used in DVB-RCS and known to both the RCST and the NCC. The information necessary to support this can be communicated in the Forward Link using the Return_Link_IV sub-type of the TRANSEC_System_Message descriptor. These initialization vectors can be used irrespective of the precise encryption technique employed. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 103 |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.1.3.3 Authentication and Key Management Hooks | It is expected that different TRANSEC profiles can have substantially different requirements and preferences for Authentication and Key Management, e.g. some Governmental profiles may only support manual key entry (using approved key fill devices), need minimal new signalling, whereas others (here) support over the air rekeying (and purging), add more complex signalling. Each of the TRANSEC profiles defines signalling elements for MAC layer transport of the associated messages, for example certificate exchanges and key update commands. This does not preclude other methods of transport of such messages, if allowed and/or preferred by specific implementations. In general, it is expected that authentication and mutual trust is established using standardized PKI/X.509 certificate solicitation and exchange. In the forward link, authentication and key management messages can be transported in system specific variants of the TRANSEC_System_Message descriptor. System specific variants are identified by certain values of the descriptor_type field. In the return link, authentication and key management messages can be transported using a specific new protocol type. This protocol type can be used with both RLE and GSE encapsulation. The content of PDU's transported with this protocol type is system specific. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.1.3.4 Traffic Flow Obfuscation Hooks | On the forward link, traffic activity should be concealed / obfuscated by transmitting dummy (or "chaff") packets in broadcast mode (i.e. with no specific terminal address). A special chaff protocol type is defined for this purpose. The dummy packets should be inserted in a frame when not enough actual data are available to fill up the frame. They should include random data which should be encrypted as actual data. On the return link the traffic activity should be concealed / obfuscated by transmitting dummy bursts. The dummy bursts should be transmitted in allocated slots, when there is no actual (useful) data to transmit. The payload of the dummy bursts should include random data, encrypted as actual data. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2 Consumer Profile | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.1 Introduction | The aim of the Consumer profile is to provide confidentiality to the higher layer traffic sent over satellite and to protect against satellite link and subscription piracy. The higher layer protocol TRANSEC provides: • Data protection: The traffic is encrypted using AES-256, with a traffic key shared by a group of RCSTs. This key is distributed over the air. • Key protection: RCST group's keys are protected in the distribution by encryption by use of a passphrase (based on a password shared between each RCST and the key manager). Encryption of the signalling system is not done in the Consumer profile. Keeping the signalling in the clear eases the installation of the RCSTs, and enables the operator to more easily do e.g. the line-up of the RCST without the knowledge of the RCST's password. The Consumer profile addresses the general user data security concern for both private and shared/public teleports. This requires offering to the users the option of having the layer 3 PDU encrypted as it is in transit in the DVB-RCS network. Different "user groups" on the network (which may be different organizations entirely, each with their own sets of RCSTs), should be allowed their private encryption keys. The Consumer profile should also support protection for the DVB-RCS network operator, e.g.: • Prevent use of subscriptions by others than the legitimate subscriber • Prevent unauthorized use of RCST hardware • Thwarting productive use of stolen or "hacked" RCSTs ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 104 • Detection and thwarting use of cloned RCST MAC addresses Thwarting denial of service attacks, with respect to network log-on. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.2 Security Architecture | The layered model for the Consumer profile is shown in Figure 8.1. The Encryption hooks are shown in red. In particular, on the forward link, performing layer 2 encryption of the unicast GSE PDU, and on the return link, performing layer 1 encryption of the exclusive traffic slots. User applications System control User applications PDU (e.g. IP) PDU (e.g. IP) Layer 2 Unicast (GSE PDU) Layer 2 Broadcast, Multicast & Signalling (GSE PDU) Layer 2 (RCS2 PDU) Layer 1 (BB Frame) Signalling slot Contention TRF slot Exclusive TRF Slot DVB-S2 TDM RCS MF-TDMA Figure 8.1: Consumer Profile Layered Security Model |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.3 Authentication and Key Management | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.3.0 Overview | Key management relations between the Feeder, Gateway and RCST are shown in Figure 8.2. Figure 8.2: Key Management Relationships |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.3.1 Security Enabling | • Determine a password for each RCST • Determine the FL unicast key group of each RCST Generates: Table of keys RCST group – keys relations Key Management Receives: Table of keys Routes - keys l ti Feeder Receives: Table of keys RCST-keys relations Gateway Receives: Table of keys Gets entered: Password RCST Encrypted Traffic ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 105 • Determine the FL multicast key group(s) of each RCST • Determine the mesh key group of each mesh RCST • Generate a traffic key for each key group • Prepare an encrypted version of the applicable traffic keys for each RCST based on the password • Enable Feeder and Gateway with the necessary key group keys • Enter password and switch on TRANSEC operation at each RCST • Switch on TRANSEC forwarding to all the RCSTs concerned • Switch on TRANSEC for all the multicast concerned • Logoff and logon all the RCSTs, so all relevant traffic keys are loaded |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.3.2 Security Disabling | Switch off TRANSEC for the Feeder. Switch off TRANSEC for each RCST. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.3.3 Passphrase generation for RCSTs | The passphrase is set by the user. It is used to encrypt the traffic key for the RCST provided at logon. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.3.4 Key renewing | Provision the new key at Feeder, Gateway, and in encrypted versions at NCC, and have the RCSTs logon again. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.3.5 Key encryption | The key encryption is done using AES-256 in CBC mode [i.21] and with the initialization vector is according to clause 8.2.4.5.2.2. The key used for encrypting the traffic key is generated by using the passphrase. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.3.6 RCST key generation | The key used for encrypting traffic keys for a particular RCST is generated from the RCST's passphrase using a predefined algorithm. The algorithm is assumed agreed between the customer and the equipment vendor. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.3.7 Key ID | The key ID is an index referring to a specific traffic key. A traffic key can be identified by its key ID. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.3.8 Key Storage in the RCST | The RCST acquires the traffic keys at logon and is not expected to store the keys when powered off. It should purge all previous traffic keys when logging on. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.4 Encryption | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.4.1 TRANSEC M&C | The TRANSEC M&C interface should only be accessible via a locally connected console or via a non-detachable panel, and not via the IP network. This will prevent illegitimate users from connecting via the IP network. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 106 Set TRANSEC password This function enters the TRANSEC password for non-volatile storage of data to retrieve the deducted passphrase. Display TRANSEC status This function should display the current TRANSEC status. The TRANSEC state should be clearly indicated to the user in a consistent and permanent manner. Clear TRANSEC password and passphrase This function clears a user-entered password and the associated passphrase. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.4.2 TRANSEC Procedures | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.4.2.1 RCST Logon | The logon procedure is the same as for the non-TRANSEC system, excepted that the RCST will also retrieve its traffic keys as part of the procedure. Upon logon, the traffic keys for the RCST, encrypted by use of the RCST's passphrase, is sent to the RCST via a dedicated descriptor in the unicast TIM. This TIM descriptor is further described in clause 8.2.4.5.1. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.4.2.2 Transmission of encrypted traffic to the Gateway | A TRANSEC enabled RCST encrypts higher layer user packets before transmission. An Initialization Vector is generated, following the Initialization vector format defined in clause 8.2.4.5.2.1. The RCST encrypts the payload using the Initialization Vector and the current key, and AES-256 in CTR mode [i.21]. A part of the IV is included in a header extension in the transmitted layer 2 payload, specified in clause 8.2.4.5.3. The payload is associated with the protocol type of the header extension for higher layer encryption. An explicit indication of this protocol type is included in the payload if the associated default protocol type is not equal to this protocol type. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.4.2.3 Reception of encrypted traffic from the Feeder | A TRANSEC enabled RCST detects the received encrypted higher layer packets and de-crypts before forwarding to the user, using the key explicitly referred to by the Key ID in the extension header, or the default key for the forward link if the Key ID is not indicated. The decryption is equal to the encryption done by the Feeder as described in clause 8.2.4.3.1.2. After removal of the extension header and decryption the RCST forwards the clear-text packet to the higher layer. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.4.2.4 Decryption of the encrypted traffic key | To decrypt an encrypted key, the RCST uses its own key encryption key that is generated by using the passphrase, and the Initialization Vector associated with the encrypted key. Decryption is equal to the encryption described in clause 8.2.3.5. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.4.3 TRANSEC by the Feeder | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.4.3.1 Procedures | 8.2.4.3.1.1 Transmission of a unicast packet If TRANSEC is enabled for the RCST that the unicast packet is aimed for, the Feeder encrypts the higher layer packet before transmission to this RCST. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 107 8.2.4.3.1.2 Encryption of a unicast packet The Feeder uses an appropriate group traffic key associated with the destination RCST. It generates and uses an Initialization vector following the Initialization vector format defined in clause 8.2.4.5.2.1. The packet is encrypted using the appropriate Initialization Vector and the unicast traffic key of the RCST, and uses AES-256 in CTR mode [i.21]. A part of the IV is included in a header extension in the transmitted layer 2 payload, specified in clause 8.2.4.5.3. The Feeder will include the Key ID in the header extension when another key than the default is used, and may omit the Key ID if the default key is used. 8.2.4.3.1.3 Transmission of a multicast packet If TRANSEC is enabled for the multicast group, the Feeder encrypts the higher layer packet before transmission. 8.2.4.3.1.4 Encryption of a multicast packet The Feeder uses an appropriate group traffic key associated with the multicast group. It generates and uses an Initialization vector following the Initialization vector format defined in clause 8.2.4.5.2.1. The packet is encrypted using the appropriate Initialization Vector and the applicable key, and uses AES-256 in CTR mode [i.21]. A part of the IV is included in a header extension in the transmitted layer 2 payload, specified in clause 8.2.4.5.3. The Feeder will include the Key ID in the header extension when another key than the default is used, and may omit the Key ID if the default key is used. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.4.4 TRANSEC by the Gateway | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.4.4.1 Procedures | 8.2.4.4.1.1 Reception of an encrypted packet Upon reception of an encrypted traffic packet, the Gateway detects that the packet is encrypted, and decrypts it with the appropriate key before forwarding it to the higher layer. 8.2.4.4.1.2 Decryption of a packet The Gateway selects the applicable traffic key for each encrypted packet received. The traffic key is the one associated with the RCST. This is a key administratively associated to the RCST that sent the packet. This RCST is identified by the burst time plan. The decryption procedure is identical to the encryption procedure used by the RCST as specified in clause 8.2.4.2.2. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.4.5 Format and syntax of TRANSEC elements | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.4.5.1 Providing the traffic key in TIM-U | The tag used for this descriptor is recommended to be 0x21. The descriptor has the following format. Table 8.1: Key Descriptor used in TIM-U Syntax Value Number of Bytes Information Mnemonic Traffic_key_descriptor() { Tag 0x21 1 Uimsbf Length 0x27 1 Uimsbf Key_use 1 Bslsbf Random_field 4 Uimsbf Key_id 2 Uimsbf Encrypted_key 32 Uimsbf } ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 108 Description of parameters: Tag 0x21 Length 0x27 Key_use control bits defined in Table 8.2 Random_field 32 bit random variable used as part of IV Key_id 16 bit key identifier Encrypted_key 256 bit encrypted traffic key Table 8.2: Administrative Encryption Context Control Bit Control Flag Encoding 0 Forward_link_default 1=yes, 0=no 1 Return_link_use 1=yes, 0=no 2 Mesh_link_use 1=yes, 0=no 3-7 Reserved The control flags indicate the contexts where a key may be used. If there are several keys allowed used for the same transmission context, the RCST is free to choose which to use. One key may be used for all the crypto contexts. TIM-U can contain several keys. If ambiguity is to be avoided, the RCST can be given a single key for each transmission context. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.4.5.2 Initialization Vectors | 8.2.4.5.2.1 IV used for traffic encryption The initialization Vectors used for traffic encryption (both on RL and FL) are generated with the following format: Date of the Day + base of the NCR value. This ensures that the IVs will be changed for every transmitted packet. Table 8.3: Initialization Vector for encryption of traffic Syntax Number of Bytes Information Mnemonic TrafficEncryptionInitializationVector() { Year 2 Uimsbf Month 1 Uimsbf Day 1 Uimsbf NCR_base 4 Uimsbf Group_id 1 Uimsbf Logon_id 2 Uimsbf Zero-Padding 5 '0' } Description of the parameters: Year Current year of the operating system encoded in a 16 bit field Month Current month of the operating system (1-12) encoded in an 8 bit field Day Current day of the operating system (1-31) encoded in an 8 bit field NCR_base 32 least significant bits of the local NCR base value sampled just in time before each encryption Group_id The assigned Group ID of the RCST Logon_id The assigned Logon ID of the RCST ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 109 At the receiver side, the NCR_base value used at encryption is reconstructed by getting the NCR_base LSBs that the transmit side has appended to the encrypted packet, take its own NCR_base value, and then recover the full NCR_base that the transmit side did use by applying the following algorithm: • Sample the local NCR_base and compare the MSB of the received NCR_base subsection with the locally generated counterpart. • If the local one is the lowest, then the subsection of the local NCR_base has wrapped relative to the one received with the packet. Subtract 2^(the number of bits of the NCR_base subsection transmitted +1) from the local NCR_base sample, and then replace the LSBs by the received NCR part. • Else, just replace the LSBs of the local NCR_base sample by the NCR subsection received with the packet. Similar processing applies as well for Date if the Date value indicates wrapping, and may then also apply for Month and Year. In-band signalling information is used to resolve synchronization issues between the hub and RCST when time-of-day clocks are slightly out of synchronization. 8.2.4.5.2.2 IV used for key encryption The initialization vector used for key encryption is a 128 bits sequence generated following the structure below. Table 8.4: Initialization Vector for encryption of a traffic key Syntax Number of Bytes Information Mnemonic KeyEncryptionInitializationVector(){ Key_id 2 Uimsbf Group_id 1 Uimsbf Logon_id 2 Uimsbf Random_field 4 Uimsbf Zero-Padding 7 '0' } Description: Key_id 16 bit key identifier Group_id The assigned Group ID of the RCST Logon_id The assigned Logon ID of the RCST Random_field As transmitted in the traffic key descriptor The IV is renewed for each key encrypted, which means that each time the NMS encrypts a new key it also generates and uses a new IV (in effect a new random part and a new key_id for a given RCST). |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.4.5.3 Crypto Context Extension Header | A new Protocol_Type (2B) value for GSE indicates that there is an Explicit_Crypto_Context_Extension header appearing first in the FL payload, followed by an encrypted packet of the indicated protocol type. Table 8.5: Crypto context extension header applied when using explicit key selection Syntax Number of Bytes Information Mnemonic Explicit_Crypto_Context_Extension(){ Key_id 2 Uimsbf Encryption_Context 3 Uimsbf Compressed_Protocol_Type 1 Uimsbf } ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 110 Description of the parameters: Key_id This 2 byte field indicates the Key ID of the key used to encrypt the packet following the extension header. Encryption_Context This 3 byte field contains for the FL the 5 LSB bits of Day as MSB and then 19 LSBs of the NCR subsection used in the Initialization Vector as LSBs. This 3 byte field contains for the RL the 24 LSBs of the NCR subsection used in the Initialization Vector. Compressed_Protocol_Type Compressed protocol type value for the clear-text variant of the encrypted PDU, as specified in ETSI EN 301 545-2 [i.1]. The encrypted packet is appended following this extension header. In the return link and on mesh links, a smaller header extension is always used, as the key ID is determined implicitly. A new Compressed_Protocol_Type (1B) value for RLE indicates the presence of the Implicit_Crypto_Context_Extension. This smaller header extension may also be used in the forward link when encrypting with the key indicated to be the default for the forward link. In this case, a further new Protocol_Type (2B) for GSE indicates the presence. Table 8.6: Crypto context extension header when using implicit key selection Syntax Number of Bytes Information Mnemonic Implicit_Crypto_Context_Extension(){ Encryption_Context 3 Uimsbf Compressed_Protocol_Type 1 Uimsbf } Encryption_Context This 3 byte field contains for the FL the 5 LSB bits of Day as MSB and then 19 LSBs of the NCR subsection used in the Initialization Vector as LSBs. This 3 byte field contains for the RL the 24 LSBs of the NCR subsection used in the Initialization Vector. Compressed_Protocol_Type Compressed protocol type value for the clear-text variant of the encrypted PDU, as specified in ETSI EN 301 545-2 [i.1]. The encrypted packet is appended following this extension header. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.2.5 Traffic Obfuscation | The Consumer profile could use DVB-RCS2 Free Capacity Assignment (FCA) capability to ensure that return link carriers and timeslots are maximally filled, in order to prevent traffic analysis based on network activity. Use of FCA ensures that RCSTs will have access to all network capacity at all times to send any available traffic. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3 Professional Profile | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.1 Introduction | The Professional profile is for markets where comprehensive, strong network security is a primary concern. It is expected that DVB-RCS networks in these markets will be private networks (with a private NCC/hub system) where the same high grade of security is provided throughout the network. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 111 Requirements for the Professional profile include: • Obfuscation to prevent traffic pattern or activity analysis on Forward link carriers, which implies the use of: - Encryption of all Layer 2, Layer 3 and higher layer information (include all related headers in transparent and mesh overlay systems. In mesh regenerative some headers or specific field may be in clear for OBP routing). - Filling of any unoccupied DVB-S2 frames with "chaff". - Encryption of all network signalling (excepting the NCR). - The option of operator choice regarding the level of traffic obfuscation on return link for mesh systems. • Encryption of payload in each burst (including signalling bursts), and any continuous transmission on these same carriers, protecting all Layer 2, Layer 3 and higher layer information. • Strong authentication: - of the Hub/NCC (to the user); - of the user (to the Hub/NCC); - between users (in mesh systems). |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.2 Security Architecture | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.2.0 General | Figure 8.3 shows the security components of the Professional profile network. Figure 8.3: Professional Profile Security Architecture Within the Indoor Unit (IDU) of each RCST there should be security functions of encryption, decryption, authentication and related security management tasks that all are built to a sufficiently qualified trust level. An isolated implementation of these functions in a qualified security module should exist, and here in a module called the RSM (for RCST Security Module). This module may be built-in at the factory, or provided as a user plug-in option for a more generic HW platform. At the Hub one or more similar hardware modules should exist, here called HSM (for Hub Security Module). Typically, one HSM may be required for the Forward Link and another for the Return Link. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 112 For operation on the network, each RSM and each HSM is equipped with its own X.509 Private Key User Certificate, plus the corresponding Root CA Public Key Certificate for the issuing Certificate Authority (CA). Each RSM and each HSM also has IPsec Tunnel Mode capability and mutual authentication capability. Each RSM and each HSM has local I/O on the module for secure management and operation. This I/O should allow the X.509 Certificates to be conveniently installed, the entry of usernames and passphrases, and other security-related management task that should be done locally. At the Hub site, there is an IPsec Tunnel Gateway which has a corresponding qualified trust level. That can be a standard commercially available piece of equipment. This device should also have an X.509 User Certificate installed with the corresponding Root Certificate from the issuing CA. The same CA is assumed used to issue the certificates of all these modules. Also, it is likely these modules will have to be compliant with national government standards on information security. Behind the IPsec Tunnel Gateway resides the Network Security Controller (NSC). This administrative module, based typically on a standard server, controls and supervises the security related status of each user terminal and electronically distributes the traffic keys required for use of the Protected Channel. Remote access to NSC is only possible via IPsec Tunnels from authorized security modules (RSM and HSM), which serves to authenticate the identity of the device at the end of the IPsec tunnel, and also encrypts the communications for privacy and integrity, plus there is protection against replay attacks. Finally there is the CA itself. It may reside at the Hub site, or elsewhere, and may be accessed via any acceptable method (electronic or otherwise) per policies of the user organization. More details on the functions of each module are provides in Table 8.7. Table 8.7: Description of Modules for the High Grade Security profile Security Component Description RCST Security Module (RSM) This is a module with a qualified trust level. It has IPsec tunnel capability for authentication and encryption. It intercepts Layer 2 control signalling. It controls the forwarding of traffic to the internal router with respect to chosen security policy and current security state. It communicates with the Network Security Controller via IPsec tunnels, to decides the current security state of the RCST. It has dedicated local I/O separated from the other parts of the RCST. It locally manages the use of traffic keys for access to the Protected Channel (see next clause) which are distributed to it by the Network Security Controller. It synthesizes the traffic key for the Acquisition Channel from external local input from Security Administrator. Hub Security Module(s) (HSMs) Similar to RSM, but on the hub side. Typically one per Forward Link carrier and one for a group of carriers supporting the Return Link. IPsec Tunnel Gateway This is a module with a qualified trust level. It restricts the communication to/from the Network Security Controller with the RSMs and HSMs to be via mutually authenticated IPsec tunnels, preventing spoofing and replay attacks. Network Security Controller This is an administrative unit controlling and supervising security via the RSMs and HSMs. It distributes the necessary crypto material (e.g. the symmetric keys with their applicable cryptoperiods) for the Protected Channel between the RSMs and HSMs. It also implements roll-over control for these traffic keys, and removes lost or compromised entities from the key distribution. CA and Key Generator This is a system with a qualified trust level, required for the provisioning of X.509 Certificates. It may also generate all traffic keys used in the network in addition to the conveyor keys (KEK). |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.2.1 Encrypted Communications Channels | Symmetrically encrypted communications of user traffic and signalling occurs between the RSM and HSM modules. This is accomplished using two duplex communications channels called: the Acquisition Channel and the Protected Channel. Their purposes, and the reason for using two such channels, are as follows: • Acquisition Channel: For initial acquisition of the Forward Link by RCST, plus terminal log-on, authentication and key distribution, as may generally be required prior to accessing the Protected Channel. • Protected Channel: For protected transport of user traffic and for RCST-NCC communications in mesh systems. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 113 In general, higher protection is provided to the Protected Channel. Breach in the Acquisition Channel encryption should not give access to the Protected Channel. Traffic keys for access to the Protected Channel are provided via the authenticated IPsec tunnel endpoints, and the keys may also be encrypted end-to-end from the Key Generator to the security module. On the Forward Link, each of these channels is implemented as one input stream in a multi-stream DVB-S2 TDM. On the Return Link both channels have a quantity of TDMA timeslots associated with them (of either variable or fixed capacity). For the Protected Channel, there may also be Mesh Link capacity on TDMA carriers. The reasons for encrypting the Acquisition Channel are: • Authenticate users attaching to the network before providing the protected mode traffic keys, which are issued "just-in-time". • Conceal information about network identity as may be revealed in SI table content sent over the Acquisition Channel. • Extend protection against traffic pattern and activity detection to the Acquisition Channel. The Professional profile requires RCSTs to support encryption on the Acquisition Channel. Some operators may, by policy, choose not to encrypt it, but they would have to ensure that the RCSTs (with their RSMs) selected for their network actually allow for a non-encrypted Acquisition Channel. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.2.2 Layered View of Architecture | Figure 8.4 shows the layered architecture used for the Professional profile. The red blocks indicate where the encryption is applied (i.e. the encryption hooks). User applications Security Mgmt & Control (with Key Management) User applications User PDUs (e.g. IP) IP User PDUs (e.g. IP) IPsec Tunnel Layer 2 Unicast & Broadcast Signaling & User Traffic (GSE PDU) Layer 2 Unicasts & Broadcasts Signaling & User Traffic (RLE PDU) Protected Channel Frame Acquisition Channel Frame Acquisition Channel Payload Protected Channel Payload DVB-S2 TDM RCS2 MF-TDMA Forward Link Carriers Return Link Carriers (and Mesh for Protected Channel) Figure 8.4: Layer Architecture for the Professional Profile Obviously, when encryption is applied at a lower layer it also encrypts the upper layers using the lower layer. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.2.3 Network Clock Reference (NCR) | The NCR is sent as a broadcast in the clear. This is done so that the NCR solution can be the same as for the non-secured DVB-RCS implementation, and so that the broadcast NCR can function as the basis of the counter in CTR mode encryption (as described in clause 8.3.4.6). The NCR alone provides no useful information about the network. No other system information is transmitted in the clear. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.3 Authentication and Key Management | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.3.1 Network Operation | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.3.1.0 Introdcution | This clause describes the operational dependencies of a network with the Professional profile, in temporal fashion from initial set-up to routine use of the Protected Channel with key roll-over. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 114 |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.3.1.1 Installation of X.509 Certificates | Before the RCST can enter the network, the RCST Security Module (RSM) should be loaded with: • X.509 User Certificate issued by a Certificate Authority (CA) • The CA's Root Certificate The User Certificate should be unique for each RCST. The Hub Security Modules (HSM) also should be loaded with their unique User Certificates and the associated CA Root Certificate. These certificates are loaded the security management port on the RSM and HSM. Also, the public keys of all RSM should be loaded into the NMC/NCC for each terminal, prior to terminal activation on the network, to support logon authentication. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.3.1.2 Security System Time of Day | The security system Time of Day (TOD) should be set by the operator. This is required to align the symmetric encryption. The accuracy should be according to system policies, but a TOD within a couple of minutes should be sufficient to enter the network. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.3.1.3 Network PASSPHRASE | There is a network PASSPHRASE, common to the network as a whole. It is used to restrict access to the Acquisition Channel. It should be known by the RCST operator and should be entered into the RSM of the RCST to gain access to the network. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.3.1.4 USERNAME | Each RCST and each HSM in the operational network is associated with a unique USERNAME (within the organization) for this installation. The USERNAME is associated with a compatible hardware unit by loading the credentials consisting of at least the user certificate and the network PASSPHRASE. The USERNAME identifies the user that has been given these credentials. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.3.1.5 First Power-on of RCST after Installation | When a newly installed RCST is powered-on, it tunes to the assigned DVB-S2 carrier. Then, using standard Base-Band Header information (which remains in the clear), the RCST, by default, selects the stream corresponding to the Acquisition Channel, where it finds the NCR (in the clear) to reach the state of "Ready for Logon". From this state the RCST could begin the log-on process. However, prior to any attempts by the RCST to log-on (i.e. prior to sending any logon bursts), the operator should have entered: • The USERNAME of the installation • The network PASSPHRASE The network PASSPHRASE, the NCR and the return link structure are used to generate the decryption key and the encryption key for the Acquisition Channel. After decrypting the Acquisition Channel the RCST processes the SI table information. From this information the RCST discovers where to find the carriers for the Return Link and all necessary structure information for this channel. Among this information the RCST learns how to submit a logon burst to the NCC, in a manner compliant with the network policy. This policy may disallow the use of Slotted Aloha access for logon bursts, and instead have exclusive logon timeslots assigned to each RCST. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.3.1.6 Content and Security of Log-on Information | The RCST forms the content of the control burst, which includes current standard elements, such as its 6-byte hardware ID, plus the following additional information required for the Professional profile: • 4 LSB bytes of the SHA-1 hash of USERNAME (of the installation); • A Logon Signature for the control burst. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 115 Every logon request should have a Logon Signature. The Logon Signature should be a chosen minimum number of bytes, taken from the least significant bytes of a Digital Signature. The Digital Signature is a standard type according to public key method (e.g. the one from IETF RFC 4880 [i.23]), signed using the private key of the RCST RSM's X.509 Certificate, and is used to sign the following concatenated information: • The 6-byte Hardware ID of the RCST • The USERNAME (of the installation) • A "Slot Position Identifier" in the context of an extended superframe counter An extended superframe counter is required, e.g. as elaborated in the Governmental profile, clause 8.4.4.2.3. The NCC/NMC can verify the authenticity of the logon and the user by checking the Logon Signature using the known public key of the RSM. This allows the NCC to ignore logon requests that are unauthorized. By using a time-variant part in the signature the logon process is protected against replay. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.3.1.7 Synthesizing the Key | The Acquisition Channel traffic key is generated at the RCST based on the network PASSPHRASE. The 256 bit key needed is synthesized using the lower 32 bytes of the SHA-1 hash of the PASSPHRASE, combined with a well-known Initialization Vector (IV). |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.3.1.8 Set-up of IPsec Tunnels between RSM and the IPsec Gateway | After logon via the Acquisition Channel, the RSM establishes an IPsec tunnel to the IPsec GW and a connection to the NSC via this IPSec GW. The RSM and IPsec GW require mutual X.509 certificate based authentication to set up the tunnel, so the IPsec GW is authenticated to the RSM and vice versa. This is done according to standard procedures specified for [i.21]. The RCST takes the initiative to set up the connection to the NSC whenever this connection is missing. It stays up all the time during the session. It is not disconnected when switching from the AC to the PC. It should be noted that the aim of setting up the security association at this layer is to allow use of a general purpose certified implementation of an IPsec gateway to isolate a security controller implementation on the hub side, with less strict requirements for qualified trust. The RSM should intercept and block the traffic when operating in the unprotected mode. It should act as the endpoint of the IPsec tunnel. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.3.1.9 Traffic Key Distribution | In transparent systems, when the secured and authenticated Layer 3 connection is established, the NSC loads the RSM with the Tx/Rx key pair needed for protected mode operation. In mesh systems, this permits the start of DCP logon with the NCC procedure, through the protected channel. The key pair and metadata may be transported in one of several different transport modes: • In the clear (the layer 3 tunnel is secured) • Asymmetrically encrypted by the CA private key • Double encrypted, by CA private key and user public key • Symmetrically encrypted by a KEK The key pair is issued with a version number. In mesh systems, the RCST to RCST dynamic link establishment procedure also should provide a Tx/Rx key pair. The same transport modes are available for mesh dynamic links through the RCST to RCST protected channel. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.3.1.10 Set-up of IPsec Tunnels between HSM and the IPsec Gateway | The HSM takes the initiative to set up a secured connection to the security controller, just as the RSM does after having logged on to the network. ETSI ETSI TR 101 545-4 V1.2.1 (2026-01) 116 |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.3.1.11 Traffic Key Distribution to the HSM from the Security Controller | The HSM is configured with the PASSPHRASE and synthesizes the traffic key for the Acquisition Channel just as the RSM does. The HSM receives the traffic keys for the Protected Channel via the IPsec tunnel just as the RSM does. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.3.1.12 Transition to Protected Channel | Once the RCST has been provided with valid keys for the Protected Channel it may attempt to logon to that channel. The logon is constructed as the logon used in the Acquisition Channel. The RSM will keep the security association with the NSC via the IPsec GW during and after the transition. The RCST should resort to logon via the Acquisition Channel whenever it does not have the current traffic keys for Protected Channel. If it has valid keys, it may attempt to acquire the Protected Channel directly. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.3.1.13 Key Roll-over on Protected Channel | Keys rollover happens in advance of expiration of the current traffic keys. It is seamless for those already attached to the Protected Channel. The keys are distributed via the IPsec Tunnels in unicast mode to each RSM currently on the Protected Channel, and each active HSM. The NSC loads future versions of the traffic key (or key pair when splitting Tx & Rx) in advance to support key rollover. These keys are provided by the Key Generator (KG). Key rollover for the Forward Link is driven at the discretion of the NSC. The key version to be used in the return link may track the version used in the forward link, similar to that proposed in the Government Profile in clause 8.4.3.2. Alternatively, more precise rollover may be done by explicit instruction unicasted to each RCST. The first half of the CRC32 field is used for an encrypted CRC16 taken over the encrypted payload, and the second half is used for a CRC16 taken over the cleartext information, including all cleartext headers, to protect against link errors. This split protects the router functions against link errors as well as the random errors that occur from the use of incorrect decryption keys. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.4 Encryption | |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.4.1 DVB-S2 Physical Layer | The DVB-S2 Physical Layer (PL) frames and PL headers should be transmitted in-the-clear to support the demodulator. |
cb3dadab4f22493b63142dd0ebb828dd | 101 545-4 | 8.3.4.2 Encryption BBFRAME Payload | Most of the baseband frame (BBFRAME) payload will be encrypted with a symmetric algorithm. There are some exceptions: • The BBFRAME header • The whole NCR frame • The Crypto Block Header (e.g. 1 byte) described in clause 8.3.4.6.1 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.