COMMUNICATION

Field of the Invention

[0001 ] The invention generally relates to satellite communication, in particular to multibeam satellite communication. More specifically, the invention relates to a method usable in the field of high-capacity broadband multibeam satellite communication and a satellite transponder for carrying out such a method. An aspect of the invention deals with a telecommunication signal for civil-use digital communication over satellite, e.g. digital video broadcasting, broadband interactive services, etc.

Background of the Invention

[0002] Current broadband multibeam satellite communication systems employ multiple gateways (also known as“gateway stations”) in the transmitting segment. Indeed, this architecture exploits the multiplexing diversity by reusing all the available feeder link bandwidth across the different gateways. In such architecture, some particular advantages can be realized, such as optimization of the feeder link resources (i.e. the bidirectional links between the gateways and the satellite), reduction of the signal processing complexity, increased reliability and backhauling cost reduction. In the user segment, spot-beam (hereinafter: multibeam) technology is used, which is characterized by multiple relatively narrowly focused spot beams serving the users in different geographical areas. Except for the areas in which adjacent beams overlap, only one of the beams is dedicated to a given spot. The multibeam technology enables frequency re-use across the multiple user beams, which together make up the footprint of the satellite. Recent studies in such architectures have shown that full re-use pattern (i.e. duplicating the same frequency resource within adjacent beams) along with a precoding technique, capable to handle inter-beam interference, provides acceptable performance.

[0003] Regarding optimization of the feeder link resources, i.e. the bidirectional links between the gateways and the satellite, this may be illustrated by the following considerations. Let us suppose there are N on-board feeds/antennas embarked on a satellite to serve a multibeam coverage area. Assuming further full frequency reuse among user beams and a single gateway in the transmitting segment, the required feeder link resources can be calculated as:

where are the required feeder link and beam bandwidths,

respectively. Any improvement of the bandwidth available to the users, i.e. of the beam bandwidth requires that the feeder link resources, i.e. are

increased accordingly. The feeder link might thus become the communication bottleneck. When a set of F gateways is used in the transmitting segment the required bandwidth in each feeder link can be reduced to:

[0004] Therefore, a multibeam system with multiple gateways in the transmitting segment supports higher bandwidth compared to the same multibeam system with only a single gateway.

[0005] The signal processing complexity can be reduced by making each gateway handle a smaller number of beams, i.e., a cluster of adjacent beams, at the satellite coverage. The signal processing on the transmitted traffic streams is thereby distributed over different gateways, leading to less complexity.

[0006] In the event of a gateway failure, traffic can be rerouted through other gateways to avoid service outage.

[0007] Finally, satellite service providers such as TV operators or Internet providers find a valuable set of gateways which are geographically distributed, so they can use a gateway that is close to their premises.

[0008] One distinguishes generally between two different multibeam architectures with multiple gateways. The first implements so-called ground-based processing (GbP), while the second uses space-based processing (SbP). These concepts are detailed hereinafter.

[0009] With GbP, the gateways precode the transmitted signals/data traffic so that each gateway computes a part of the full precoding scheme in the transmitting segment. The payload of the satellite works in transparent mode, i.e. without performing signal processing on the transmitted streams. It has been shown in V. Joroughi, M. A. Vazquez, and A. I. Prezez-Neira “Precoding in Multigateway Multibeam Satellite Systems,” in IEEE Transactions on Wireless Communications, Vol. 15, pp. 4944-4956, July 2017, and in G. Zheng, S. Chatzinotas, and B. Ottersten, “Multi-gateway cooperation in multibeam satellite systems,” in Proc. IEEE 23rd Int. Symposium Personal Indoor Mobile Radio Commun. (PIMRC), Sep. 2012, pp. 1360- 1364, that developing precoding in multibeam systems with multiple gateway entails taking into account several issues. First, each gateway serves a cluster of beams. Second, since the data traffic is independently generated at each gateway, every gateway must precode the signals in a decentralized fashion and transmit them through their corresponding feeder link, one feeder link per gateway. In other words, the overall precoding scheme is computed in a distributed manner at the gateways so that each gateway can only use certain feed elements. Third, the performance of the precoding scheme is intuitively sensitive to the user and feeder link Channel State Information (CSI) qualities at each gateway, i.e. the higher the accuracy of the CSI, the better is the performance of precoding scheme. Thus, some inter-gateway communication is required to exchange the CSI among gateways. Moreover, a CSI feedback mechanism between gateways and user terminals must be developed which is robust to the feedback and quantization errors. Forth, the employed inter-gateway link among gateways, in addition to the possibility to exchange the CSI, should be able to exchange data traffic among gateways in order to provide for the rerouting of traffic in case of a feeder link or gateway outage.

[0010] In particular, if each gateway generates individual data traffic and serves a cluster of adjacent beams in the multibeam coverage area, it has to implement an individual precoding/ scheme mitigating intra-cluster and inter-cluster interference. “Intra-cluster interference” means interference among beams belonging to the same cluster (and that are served by the same gateway).“Inter-cluster interference” means interference among beams belonging to different clusters (and that are thus served by different gateways). To cope with intra-cluster interference it is necessary that the respective gateway receives the CSI from users served by its beam cluster. To mitigate inter-cluster interference, it is necessary that the gateways of neighbouring clusters exchange CSI relating to the beams in the mutual boundaries of the clusters. This implies a highly complex transmitting segment infrastructure. [0011 ] When SbP is implemented, the satellite calculates the precoding scheme whereas signal processing at the gateways is limited to encapsulating the data streams in accordance with a high-throughput transmission standard, like DVB-S2 (ETSI EN 302 307-1 V1.4.1 (2014-11 ),“Digital Video Broadcasting (DVB); Second generation framing structure, channel coding and modulation systems for Broadcasting, Interactive Services, News Gathering and other broadband satellite applications; Part 1 : DVB-S2”) or DVB-S2X (ETSI EN 302 307-2:’’Digital Video Broadcasting (DVB); Part 2: DVB-S2 Extensions (DVB-S2X)), and to transmitting the encapsulated data streams to the satellite without applying any extra signal processing on the transmitted signals.

[0012] When using SbP, the following advantages can be realized with respect to

GbP. First, it is not necessary to establish a CSI feedback mechanism between the satellite and the gateways which leads to a shorter CSI latency. Second, no CSI exchange mechanism is needed among the gateways, leading to a less complex transmitting segment infrastructure. Third, in case of a failing gateway, the traffic can be easily rerouted to the satellite through other gateways without applying any extra signal processing at the new gateway(s).

[0013] Despite of these advantages, SbP implies that the processing burden is shifted at least in part from the gateways to the satellite payload, which represents an important obstacle for implementing this architecture.

[0014] WO 2014/001837 discloses a satellite communication system aiming at improving the achievable performance in terms of capacity when fading events affect some of the feeder links and the end-to-end link availability. The document also proposes a payload architecture optimized in terms of hardware required (to be understood in terms of number of equipment and associated power consumption, power dissipation and mass) to achieve the additional functionalities. In the event of degradation of one or more feeder links, data are switched from gateways with degraded feeder links to gateways with non-degraded feeder links using a communication network that interconnects the gateways, without that the user terminals experience service outage. To enable the switching of data from one gateways to another in case of outage, each user beam is associated to several gateways. Furthermore, the satellite payload is configured for combining individual carriers coming from different feeder links into respective user link signals. Summary of the Invention

[0015] The present invention relates to an on-board traffic-switching scheme for a broadband communication satellite that dynamically serves traffic requested at a plurality of user beams, within the power-limited environment of a communication satellite.

[0016] A first aspect of the invention relates to multibeam satellite communication method, wherein a satellite transponder:

o receives inbound telecommunication signals from a plurality of gateways, the inbound telecommunication signals carrying respective data streams relating to different services;

o transmits the data streams on a plurality of outbound RF (radiofrequency) telecommunication beams to spatially distributed users;

o while dynamically switching the data streams of different inbound telecommunication signals onto the outbound RF telecommunication beams depending on service requests emanating from the users.

[0017] This invention thus introduces signal processing onboard the satellite that allows commanding the satellite communication payload configuration in accordance with the momentary required configuration for the traffic information stream intended for multiple end users in the satellite communication coverage.

[0018] It will be appreciated that the present invention breaks the paradigm that each gateways serves data traffic to a (generally) fixed cluster of user beams and that each user beam (operating in unicast or multicast mode) receives its data traffic from a single gateway. More precisely, in the proposed architecture, each user terminal can dynamically request data traffic from any gateway. The architecture provides the flexibility to manage irregular and time-variant traffic requests within the different user beams.

[0019] Indeed, the method provides for traffic-aware routing of the data streams on board the satellite. The transponder is thus capable of flexibly switching the data streams originating from different gateways toward any of the outbound beams and thus to any of the receiving terminals in the coverage area. When data streams are generated through a set of geographically distributed gateways, the on-board data stream switching possibility is a promising technological enabler for the management of irregular and time-variable traffic requests within the different user beams. Thanks to the invention, the availability of services becomes less dependent on the geographical location of a user. This is an important step forward regarding the portability of service subscriptions (e.g. if a user moves to another location) and in terms of providing satellite communication services to moving user terminals.

[0020] As used herein, the term “service” designates an individual broadcast or interactive service (such as a TV, HDTV, audio or multimedia channel, a news service, an Internet service) destined to all or part of the users, subject or not to the users subscribing to the service in order to actually have access to it. It should be noted that each inbound telecommunication signal corresponds to a particular data stream (e.g. transport or generic stream of the DVB S2X standard) that relates to a specific set of services. A“service request” is a request from a user for a service, e.g. one that is currently not available in the user beam that serves his location.

[0021 ] The gateway infrastructure is currently cluster-oriented and implies that the availability of services primarily depends on the geographical location of the user. Thanks to the present invention, it is possible to adopt a service-oriented allocation of the gateway resources.

[0022] The feeder links between the gateways and the satellite (of which the uplink part carries the inbound signals) as well as the user links between the satellite and the user terminals (of which the downlink part carries the outbound signals) are preferably bidirectional links.

[0023] Preferably, upon receipt of a service request from a user, the transponder determines which data stream relates to the requested service, and switches the respective data stream relating to the requested service on the outbound RF telecommunication beam that serves the requesting user.

[0024] To achieve this, the transponder preferably maintains a database or lookup table indicating which data stream relates to which service and, upon receipt of a service request from a user, determines which data stream relates to the requested service using the database or lookup table. In the indication which data stream relates to which service, the data stream could be identified by a data stream identifier and/or by a signal identifier of the inbound telecommunication signal carrying the data stream and/or by a gateway identifier of the gateway from which the inbound telecommunication signal carrying the data stream is received.

[0025] Preferably, the database or lookup table indicates which outbound RF telecommunication beam serves which user, and the transponder determines on which outbound RF telecommunication beam a data stream is to be switched by using the database or lookup table. This feature is particularly advantageous in a scenario with mobile user terminals (e.g. on terrestrial vehicles, water vehicles or aircrafts).

[0026] The database or lookup table could further indicate which outbound RF telecommunication beam carries which data stream and/or which service.

[0027] The transponder is preferably configured to update the database or lookup table upon detection of a change in an association indicated therein.

[0028] The data streams carried on the inbound telecommunication signals are preferably encapsulated in container frames (e.g. so-called “super-frames”), each container frame including further an identifier (e.g. a metadata field) identifying the data stream, the inbound telecommunication signal carrying the data stream and/or the gateway from which the inbound telecommunication signal carrying the data stream was received. The identifier could also individually identify the services to which the data stream relates.

[0029] The identifier is preferably set by the respective transmitting gateway (e.g. by filling the corresponding metadata field) and allows the transponder to determine the set of services or even the individual services carried by the data stream without decoding the data stream itself. This greatly reduces the necessary processing complexity on the satellite payload. Indeed, the transponder only has to locate and decode the identifier, which may be very small in comparison with a container frame, and may forward the container frames essentially unchanged.

[0030] According to a preferred embodiment of the method, the container frames are precoded on-board the satellite before transmission on the outbound RF telecommunication beams. On-board precoding increases the computational effort on the satellite but it is advantageous in the context of the invention since there is no fixed allocation between the data streams on the inbound signals and the outbound user beams. Furthermore, implementation of a high throughput full frequency reuse scheme among the beams may imply that the inter-beam interference becomes the bottleneck of the communication when the on-board switching mechanism is deployed. Therefore, applying interference mitigation techniques, such as precoding, may become essential for efficient use of the available resources. CSI is required to carry out the precoding. The CSI could be provided to the satellite by any path. Preferably, however, the method comprises receiving the CSI from the users via return links to the satellite, because this guarantees that the most up-to-date CSI is used to perform the precoding on the satellite.

[0031 ] According to a preferred embodiment of the method, the outbound RF telecommunication beams comply with the DVB-S2X standard.

[0032] Preferably, the inbound telecommunication signals are transmitted on optical feeder links.

[0033] A second aspect of the invention relates to a satellite transponder for multibeam satellite communication. The transponder comprises:

o a plurality of feeder ports for receiving inbound telecommunication signals from a plurality of gateways, the inbound telecommunication signals carrying respective data streams relating to different services;

o a plurality of user beam ports for transmitting the data streams on a plurality of outbound RF telecommunication beams to spatially distributed users; and o a forward link section interconnecting the feeder ports with the user beam ports. The forward link section comprises a switching hub configured and arranged to dynamically switch the data streams of different inbound telecommunication signals onto the outbound RF telecommunication beams depending on service requests emanating from the users.

[0034] The satellite transponder preferably comprises one or more processors for processing the service requests, configured in such a way that, upon receipt of a service request from a user, they determine which data stream relates to the requested service, and control the switching hub in such a way that it switches the data stream relating to the requested service on the outbound RF telecommunication beam that serves the requesting user.

[0035] The satellite transponder may further comprise a memory containing a database or lookup table maintained by the one or more processors, the database or lookup table indicating which data stream relates to which service, the one or more processors being further configured to determine, upon receipt of a service request from a user, which data stream relates to the requested service using the database or lookup table.

[0036] The one or more processors are preferably configured to record in the database or lookup table which outbound RF telecommunication beam serves which user, and wherein the one or more processors are further configured to determine on which outbound RF telecommunication beam a data stream is to be switched using the database or lookup table.

[0037] A third aspect of the invention relates to a telecommunication signal comprising a physical-layer super-frame for modulating a carrier, the super-frame containing a plurality of frames with data of a data stream relating to a set of services, wherein the super-frame further comprises an identifier identifying the data stream, the telecommunication signal itself and/or a gateway sending the telecommunication signal, the identifier being inserted in the container before, between or after the frames such that it is located apart from and decodable separately from the frames.

[0038] The identifier could further individually identify the services to which the data stream of the super-frame relates.

[0039] A third aspect of the invention relates to a method of operating a gateway of a satellite communication system, wherein a telecommunication signal as described herein is transmitted.

[0040] It is worthwhile noting that the present invention is compatible with a beam hopping scheme, wherein for a given time slot, certain user beams are switched passive (no traffic transmitted) while other user beams are active (traffic transmitted), and wherein the activation or deactivation of a beam for a given time slot depends on the requested traffic (services) and the bandwidth necessary for serving the requested traffic. If the gateways were to handle the beam-hopping scheme, precise collaboration of the gateways would be required, which would only be possible through establishing an inter-gateway network, leading to a highly complex and unrealistic transmitting segment. It will thus be appreciated that thanks to the present invention, beam-hopping could be implemented and managed by the satellite payload as a complement to the traffic-aware routing. Brief Description of the Drawings

[0041 ] The accompanying drawings illustrate several aspects of the present invention and, together with the detailed description, serve to explain the principles thereof. In the drawings:

Fig. 1 : is a schematic flowchart of the communication chain of a multibeam satellite communication system employing payload-based user-initiated data traffic switching;

Fig. 2: is a diagram of a conventional physical-layer super-frame according to the DVB-S2X standard and modified super-frame containing an Origin Flag indicating which gateway prepared the super-frame;

Fig. 3: is a schematic view of a multibeam satellite communication system according to a simplified example of the invention.

Detailed Description of a Preferred Embodiment

[0042] Fig. 1 illustrates a broadband telecommunication method based upon the DVB-S2X standard that implements an embodiment of the present invention. Specifically, Fig. 1 shows the steps of the telecommunication method carried out at the transmitting ground segment (i.e. the gateways) 10, the space transmitting segment (i.e. the satellite payload) 12 and the ground receiving segment (i.e. the user terminals; also: user segment) 14.

[0043] Turning first to the ground transmitting segment, each gateway encapsulates the data streams relating to a specific set of services in DVB-S2X super-frames. These super-frames are transmitted to the satellite via a feeder link 16, which may be an RF link or, more preferably, an optical link. Each super-frame is contains an identifier, hereinafter called Origin Flag” (F), inserted in the Super-Frame Format Indicator (SFFI) field. Fig. 2 shows a comparison of a conventional super-frame (example based on DVB-S2X standard, section E.3.4) with a super-frame containing such an Origin Flag. The entire super-frame has a length of 612 540 symbols. It starts with the Start- Of-Super-Frame (SOSF) field, followed by the SFFI field and a sequence of 9 bundled PL (physical layer) frames. The super-frame is terminated by a field of dummy symbols (DUM). Each bundled PL-frame starts with a header (PLH or PLHeader) and a first pilot field (modulated pilot symbols, P2) and further contains data (Data) interleaved with second pilots (P). Each super-frame is scrambled by the gateway, scrambler resets being aligned with the start of each super-frame. Further details of the conventional super-frame structure can be found in the DVB-S2X standard, section E.3.4. The super-frame proposed in this context differs from the conventional one in that part of the space of the original SFFI field is occupied by the Origin Flag (F) and that the information of the SFFI is encoded on fewer symbols.

[0044] As an alternative to changing the structure of the SFFI field, the Origin Flag could be encoded using the 11 currently reserved format numbers of the super-frame (values 0101 - 1111 of the bsFFi vector). In this case, it would be the SFFI field itself that identifies the gateway of origin.

[0045] The Origin Flag indicates which gateway has generated the super-frame. If the allocation between the gateway and the set of services contained in its data stream is known to the satellite, the identification of the gateway is sufficient for the satellite to determine to which services the super-frame relates. The allocation between the gateways and the respective sets of services could be communicated to the satellite by any uplink signal. This could be done according to a regular or irregular schedule or in an ad-hoc manner upon a change in the set of services delivered by any of the gateways.

[0046] Turning again to Fig. 1 , the satellite payload comprises a transponder configured as a switching hub 18. The switching hub 18 comprises an inbound signalling unit performing synchronization (frequency and time) and detection of the super-frames received from different gateways.

[0047] The detection procedure includes descrambling each super-frame, detecting the SOSF and SFFI fields of each super-frame and reading the Origin Flag. In general, to detect and extract data traffic from a super-frame in DVB-S2X standard, a lot of tasks have to be carried out, on the physical layer, the MAC layer and further on the network layer. To detect and read the Origin Flag, only a very few steps need be carried out (all related to the physical layer) by the satellite payload.

[0048] To keep the computational complexity of the detection procedure at the payload on a low level, the super-frames transmitted by the gateways can be converted to baseband at the payload. The different steps of the detection procedure can then be carried out in baseband. [0049] A switch (SWu) 20 is in charge of switching the received super-frames on the different outbound RF beams. The switch 20 maintains a look-up table indicating which gateway is in charge of which service. Upon receiving a service request from a user in a given beam, the switch 20 looks up the corresponding gateway and switches the super-frames originating from that gateway onto the beam serving the requesting user. This does not exclude that there may be a default routing of some or all of the incoming super-frames, i.e. there may still be clusters of beams served by their allocated gateways. Super-frames relating to services specifically requested by users would be inserted between the default super-frames using a time- and/or frequency-multiplex transmission. Optionally, the SFFI of the outgoing super-frames could be set to a bsFFi vector compliant with the current DVB-S2X standard, because there is no need of the Origin Flag in the user terminals. On the other hand, the removal of the Origin Flag is not mandatory, since user terminals need not detect it and, therefore, the presence of the Origin Flag does not introduce additional complexity into the processing in the user terminals.

[0051 ] The transponder then precodes the super-frames in accordance with the precoding matrix. The super-frames of all user beams being aligned in time, the precoding matrix needs to be refreshed based on the CSI feedback from the user terminals only at a rate of 1 /TSF, where TSF is the duration of a super-frame. The transponder could use a conventional precoding scheme, such as zero-forcing (ZF) maximum ratio transmission (MRT), and transmit Wiener precoding.

[0052] The super-frames are transmitted to the user terminals via the RF user beams 22. Aside from receiving the data streams, the user terminals estimate the CSI, which they feedback to the satellite, and transmit service requests to the satellite.

[0053] In the remainder of this section a simple model of the aforementioned on-board switching process is presented. It is further shown how the switching process can be implemented with an on-board precoding scheme in a multibeam system.

[0054] A multibeam system considered, wherein by employing a typical Time Division Multiplexing Access (TDMA) scheme, at each time instant (slot) a set of F gateways serves a set of user terminals using K beams originating from a single satellite. To generate the beams, the satellite may be equipped a feed reflector antenna with N feeds, of which each generates a single beam, i.e. N = K. It should be noted, however, that the case N > K follows straightforwardly. In such a multibeam system with K representative users (one per beam), the signal model at multibeam coverage becomes:

where the received super frames at K user terminals are collected in a vector y = of size denoting the super frame received at the i-th user

terminal. The matrix H of size x is the channel matrix at the user link. If one assumes that the feeder link is noiseless and fully calibrated, its effect can be neglected in signal model in (3). The vector of size Kx1 is a total of K transmitted

super frames through F gateways and S _i denotes the i-th transmitted super frame. Furthermore, matrix G of size KxK is the weight of the on-board precoding and switching scheme. G can be represented as:

where G ₁ and G ₂ relate to the precoding and the switching mechanism, respectively.

[0055] A possible design of G ₁ would be, for a zero-forcing precoding scheme:

where is the Hermitian transpose of the respective matrix (i.e. the complex conjugate of the transpose matrix). The scalar a is a power factor and can be calculated as:

where P represents the total power at N feeds with (N=K).

[0056] The matrix G ₂ , representing the switching mechanism, can be written as: where vector denotes the i-th row of matrix G ₂ and is a zero matrix with one element equal to one (non-zero). Indeed, the element equal to one determines the destination of super frame s _i in (3). Therefore, the position of the single“1” in each row of G ₂ , i.e. the vector g _i , can be dynamically changed depending on the traffic streams (services) requested at each beam, leading to dynamic allocation of traffic to the user beams. Note that in (7) denotes the transpose of the corresponding matrix.

[0057] For further illustration, let us consider the multibeam satellite communication system architecture shown in Fig. 3, with K=N=3 beams (named herein B1, B2 and B3) and feeds, wherein each beam serves a single user terminal. In addition, we suppose that F=3 gateways are employed to serve the user terminals via one satellite.

[0058] At a first time instant, the matrix G ₂ may be:

meaning that super-frame S2 (from gateway number 2) is to be switched on Bi (user beam number 1 ), super-frame si goes to B2 and super-frame S3 goes to B3.

[0059] For the next time slot, the matrix G ₂ may be:

meaning that super-frame S3 is to be switched on Bi, super-frame s1 goes to B2 and super-frame S2 goes to B3.

[0060] In the event of the failure of a gateway (e.g. gateway 2), the switching matrix

G ₂ of the previous example could become:

if the missing of super-frame S2 is not otherwise taken care of.

[0061 ] While specific embodiments have been described herein in detail, those skilled in the art will appreciate that various modifications and alternatives to those details could be developed in light of the overall teachings of the disclosure. Accordingly, the particular arrangements disclosed are meant to be illustrative only and not limiting as to the scope of the invention, which is to be given the full breadth of the appended claims and any and all equivalents thereof.

[0062] In particular, it is worthwhile noting that an example of the invention was described in the context of the DVB S2X standard because the latter is widely used in current multibeam satellite communication and well understood by the persons skilled in the art. To the extent that implementation of the present invention may require changes to the standard, the skilled person will be readily aware of those changes upon reading the present document. It is, however, emphasized that the present invention is not limited to the DVB S2X standard or future developments thereof and could be used in satellite communication systems relying on other transmission standards.