Om a mrc method and system using fdm transmission

Abstract

The present invention relates to a messaging method intended for an OMA MRC telecommunication system having M sources s i iε{1,...,M], potentially L relays (r1...,r L ) and one destination. The transmission is of the FDM type and is carried out on a band divided into B mutually orthogonal subbands. The method comprises: the simultaneous transmission of M sources at a certain time interval, each source allocating at least one subband; and at least one cooperative retransmission by at least one relay node selected using a selection policy from the M sources and the L relays at a certain time interval, each selected node allocating at least one subband.

Description

TECHNICAL FIELD

[0001] The present invention relates to the field of digital communications. Within this field, the present invention more particularly relates to the transmission of encoded data between at least two sources and a destination in the case where the data is relayed by at least two nodes which can be relays or sources.

[0002] It should be understood that a relay has no message to transmit. A relay is a node dedicated to relaying messages from a source, whereas a source has its own message to transmit and in some cases this source can also relay messages from other sources, i.e. in this case the source is said to be cooperative.

[0003] There exist many relay techniques which are referred to as: “amplify-and-forward”, “decode-and-forward”, “compress-and-forward”, “non-orthogonal amplify-and-forward”, “dynamic decode-and-forward”, etc.

[0004] The present invention is particularly, but not exclusively, applicable to the transmission of data via mobile networks (for example for real-time applications) or the transmission of data via for example sensor networks.

[0005] Such a sensor network is a multi-user network which consists of a plurality of sources, a plurality of relays and a receiver using a time orthogonal multiple access scheme of the transmission channel between the relays and the destination denoted as OMAMRC (“Orthogonal Multiple-Access Multiple-Relay Channel”). BACKGROUND

[0006] An OMAMRC transmission system implementing slow link adaptation is known from application WO 2019 / 162592 published on August 29, 2019. The described OMAMRC telecommunication system has M sources, optionally L relays and one destination, M ≥ 2, L ≥ 0, and uses a time orthogonal multiple access scheme of the transmission channel which is applied between the nodes taken from the M sources and the L relays.

[0007] The maximum number of time slots per transmitted frame is M + T max where M time slots are allocated to the consecutive transmission of the M sources during a first phase and T 已用 ≤ T max time slots for one or more cooperative retransmissions are allocated to one or more nodes selected by the destination according to a selection policy during a second phase.

[0008] The considered OMAMRC transmission system comprises at least two sources, wherein each of these sources is able to operate as a source or as a relay node at different time instants. The system can optionally further comprise relays. The term "node" encompasses equally relays and sources acting as relay nodes or sources. The considered system makes it possible for a source to be itself a relay. A relay is different from a source in that it has no message to transmit that is specific to it, i.e. a relay only retransmits messages originating from other nodes. A relay always performs a cooperative retransmission.

[0009] The links between the various nodes of the system are subject to slow fading and Gaussian white noise. The destination has no knowledge of all the links of the system (CSI "Channel State Information"). Indeed, the links between sources, between relays, between relays and sources are not directly observable by the destination and the knowledge of these links by the destination would require an excessive exchange of information between sources, relays and destination. In order to limit the feedback overhead, it is assumed that the destination only knows information related to the channel distribution / statistics of all the links (CDI "Channel Distribution Information"), for example the average quality of all the links (e.g. average SNR, average SINR) to determine the rate assigned to the sources.

[0010] The link adaptation is slow, i.e. before any transmission, the destination assigns an initial rate to the sources knowing the distribution of all the channels (CDI "Channel Distribution Information"). Typically, the CDI distribution can be fed back based on the knowledge of the SNR or average SINR of each link of the system.

[0011] The message transmission from the sources is divided into frames during which the CSI of the links is assumed to be constant (slow fading assumption). The rate assignment is assumed not to change within a few hundreds of frames, it only changes with any change of the CDI.

[0012] The method distinguishes three phases, an initialization phase, and for each frame to be transmitted, a phase 1 and a phase 2. The frame is transmitted in two phases, optionally preceded by an additional phase called the initialization phase.

[0013] During the initialization phase, the destination determines an initial rate for each source by considering the average quality (e.g. SNR) of each link of the system.

[0014] The destination evaluates the quality (e.g. SNR) of the following direct links based on known techniques using reference signals: source to destination and relay to destination. For example, the quality of the source-source link, relay-relay link and source-relay link is evaluated by the source and the relay by using reference signals. The source and the relay transmit the average quality of the links to the destination. This transmission takes place before the initialization phase. Since only the average of the link quality is considered, this average is refreshed on a long time scale, i.e. on a time that allows the fast variations (fast fading) of the channel to be averaged. This time is of the order of the time required to cover several tens of wavelengths of the frequency of the transmitted signal at a given speed. The initialization phase takes place for example once every 200 to 1,000 frames. The destination feeds back to the source the initial rate it has determined via the return channel. The initial rate remains constant between two occurrences of the initialization phase.

[0015] During the first phase, the M sources transmit their messages successively during M time slots using respectively the modulation and coding scheme determined according to the initial rate. During this phase, for each source, the number of channel uses (channel use being a resource element according to the 3GPP terminology) N1 is fixed and identical.

[0016] During the second phase, the messages from the sources are retransmitted by the relay and / or the sources in cooperation. This phase lasts a maximum of T max slots. During this phase, for each source, the number of channel uses N2 is fixed and identical.

[0017] During the first phase, the sources, which are independent of each other, broadcast their information sequences encoded in the form of messages for the attention of a single recipient. Each source broadcasts its message at the initial rate. The destination sends its initial rate to each source via a control channel with very limited traffic. Thus, during the first phase, the sources in turn transmit their respective messages during the time slots dedicated respectively to one source.

[0018] The sources other than the transmission sources and the optional relay of the "half-duplex" type receive the successive messages from the sources, decode them and, if the messages are selected, generate a message based only on the source messages decoded without error.

[0019] The selected nodes then access the channel in a time-orthogonal manner to each other during the second phase so that the messages generated by these nodes are retransmitted to the destination.

[0020] The destination can select which node must retransmit at a given time instant.

[0021] This method implements a strategy in order to maximize the average spectral efficiency (utility metric) within the considered system, subject to individual Quality of Service (QoS) of each source, i.e. the average individual BLER of each source:

[0022]

[0023] where:

[0024] • R i = K i / N1represents the initial rate of source i, where K i is the number of information bits of the message from source i e {1,..., M}. R i is a variable taking discrete values from a finite set , where n MCS is a rate number corresponding to various Modulation and Coding Schemes (MCS) available for transmission;

[0025] • T 已用 ≤ T max represents the number of cooperative retransmissions used during the 2nd phase, is the average number of cooperative retransmissions used during the 2nd phase;

[0026] • a = N2 / N1is the ratio of the number of channel uses during the 2nd phase to the number of channel uses during the 1st phase;

[0027] • BLER i represents the block error rate of source i. BLER i represents a multivariate function BLER M (R1,..., R i ) depending on the current values taken by the rate variables R1,..., R M .

[0028] The QoS constraint on the individual BLER provided to each source is expressed as: The problem of optimizing the multi-dimensional rate allocation is solved using an algorithm based on the interference-free or "Genie Aided" approach. This method involves determining each initial rate of a source independently by assuming that all messages of the other sources are known at the destination and the relays, and then determining the rates iteratively by initializing the values of the rates with the values determined according to the "Genie Aided" approach. The utility metric involving the spectral efficiency depends on the strategy used to select the nodes that appear during the 2nd phase. SUMMARY

[0029] The subject of the invention is a method for transmitting messages, intended to be used for a system with M sources s ii∈{1,...,M}, optionally L relays r1,...,rL L and an OMA MRC telecommunication system with M≥2, L≥0, M≤B sources and one destination. The transmission is of the FDM type and is on a band divided into B mutually orthogonal subbands. The method is such that it comprises:

[0030] - the simultaneous transmission of the M sources during a certain time slot, each source being allocated at least one subband; and

[0031] - at least one relay node selected from the M sources and the L relays performs at least one cooperative retransmission during a certain time slot, each selected node being allocated at least one subband, in order to maximize a quality of service metric.

[0032] The allocation of subbands among the sources allows to reduce the time needed to transmit the data, since the sources transmit simultaneously in the same first time slot. Such a method is therefore well suited for services with stringent delay requirements. The allocation of one or more subbands to each source and the strategy for selecting the sources during the next time slot are performed in order to maximize a quality of service metric (e.g. BLER, spectral efficiency). Maximizing the quality of service allows to optimize the rate or to allow a reduction of the transmission power of the sources at the same rate.

[0033] One or more time slots after the first time slot are dedicated to retransmissions, including at least one cooperative retransmission. A cooperative retransmission is a transmission by a relay or a transmission by a source that is able to assist the destination in decoding at least one other source. In the following, a non-cooperative retransmission will be referred to as a retransmission of a source of its own message. A cooperative retransmission is a transmission by a node that contains information related to at least one message from another node. The transmission by a relay is intrinsically a cooperative retransmission, but also includes the transmission by a source (that is able to cooperate) that includes in its transmission information related to at least one message from another source. The cooperation of the relay nodes ensures an increased reliability of the transmission.

[0034] According to one embodiment, the selection strategy is such that a relay node that decodes a set of sources at time slot t can only cooperate at time slot t+1 for a single source of its set.

[0035] This embodiment allows to obtain a direct representation of the individual outage events of the sources, i.e. it is not necessary to obtain the outage events of all subgroups of sources that contain the considered source. The selection of the source to cooperate with can be random among the sources that have not been decoded without error by the destination, and therefore the transmission of the node to the destination includes an indication of the source that cooperated with the node. Moreover, the common outage event of a set of sources is only obtained as a joint of the individual outages of the sources in the set.

[0036] According to one embodiment, the method is such that:

[0037] - the destination broadcasts to the relay nodes its correctly decoded source set from the received sources during the transmission time slot;

[0038] - the relay nodes that have correctly decoded a source that was not correctly decoded by the destination inform the destination of this fact;

[0039] - the destination broadcasts to the relay nodes the vector a t , these relay nodes comprising the relay nodes selected for the subbands for a cooperative or non-cooperative retransmission during the next transmission time slot.

[0040] According to this scheme, the destination feeds back to the relay nodes its correctly decoded source set upon completion of the reception of the data transmitted during the transmission time slot. This feedback can occur via a control channel. According to a particularly simple embodiment, the destination feeds back M bits, these M bits indicating whether each of the M sources was correctly decoded or not. If all the sources were correctly decoded by the destination, i.e. its correctly decoded source set contains the M sources, a new frame is transmitted.

[0041] According to one embodiment, the relay nodes inform the destination by transmitting a single bit in the control channel.

[0042] According to this embodiment, the signalling from the relay nodes to the destination is minimal and thus has the advantage of consuming almost no channel resources. With this information, the destination can implement a selection policy that for example involves maximizing the sum of mutual information between the nodes that can be used to allocate the subbands and the destination at a given time slot t:

[0043] According to one embodiment, the relay nodes inform the destination by transmitting their correctly decoded source set.

[0044] Compared to the previous embodiment, the signalling from the relay nodes to the destination according to the present embodiment consumes more channel resources. However, the information transmitted allows the destination to more efficiently select the relay nodes to assist it in decoding the maximum number of sources.

[0045] According to one embodiment, the destination selects the relay nodes that allow it to correctly decode as many sources as possible upon completion of the cooperative or non-cooperative retransmission.

[0046] According to this embodiment, the destination feeds back to the relay nodes the vector in which the relay nodes are selected that maximize the number of sources that are correctly decoded by these relay nodes and that have not been correctly decoded by the destination. This vector further comprises the allocation of the subbands to the selected relay nodes.

[0047] According to one embodiment, the destination selects a relay node such that the sum of mutual information between the node and the destination, which can be assisted by the subbands allocated by the node, is maximized.

[0048] In several vectors a t In cases where the maximum number of sources that can be decoded is the same, this method selects the sum of mutual information between the node and the destination that can be assisted by its allocated subband. Maximized vector

[0049] According to one embodiment, the method features slow link adaptation and ensures that the rate allocated to the source is determined to maximize a metric expressed as an average utility function that follows an average individual BLER for each source:

[0050] in:

[0051] · Let i be a variable representing the initial rate assigned to source i, where i∈{1,...,M};

[0052] ·K i It is from source i in n 0，i ×F times the amount of data transmitted on the channel;

[0053] ·T 已用 This is the number of time slots used for cooperative retransmission / optionally non-cooperative retransmission;

[0054] · It is the average number of time slots used for cooperative retransmission / optionally non-cooperative retransmission;

[0055] ·BLER i It is the block error rate of source i.

[0056] According to one embodiment, the method features fast-link adaptation and ensures that the rate allocated to the source is determined to maximize a metric expressed as an average utility function, which is subject to a separate cutoff from the source:

[0057] in:

[0058] · It is the individual cutoff probability of source i in time slot t of cooperative retransmission / optional non-cooperative retransmission;

[0059] ·T 已用 This is the number of time slots used for cooperative retransmission / optionally non-cooperative retransmission;

[0060] · is a variable representative of the initial rate allocated to the source i, i e {1,...,M}.

[0061] Another object of the application is a system comprising M sources s1,...,s M , L relays r1,...,r L and a destination d, M≥2, L≥0, for implementing the transmission method according to the application.

[0062] Another object of the application is each specific software application on one or more information media, in which said application comprises program instructions suitable, when these applications are executed by a processor, for implementing the transmission method.

[0063] Another object of the application is configured memories comprising instruction codes corresponding respectively to each specific application.

[0064] The memories can be incorporated into any entity or device capable of storing programs. The memories can be of the ROM type (for example, CD ROM or microelectronic circuit ROM) or even of the magnetic type (for example, USB key or hard disk).

[0065] Furthermore, each specific application according to the application can be downloaded from a server accessible on a network of the Internet type.

[0066] The optional features set out above in the context of the transmission method can optionally apply to the software applications and to the aforementioned memories. BRIEF DESCRIPTION OF DRAWINGS

[0067] Other characteristics and advantages of the application will become more clearly apparent on reading the description of embodiments, given by way of simple illustrative and non-limiting examples, and the description of the drawings in which:

[0068] [ Figure 1 ] Figure 1 is a diagram of an example of an OMA MRC (Orthogonal Multiple Access Multiple Relay Channel) system according to the application;

[0069] [ Figure 2 ] Figure 2 is a diagram of a transmission period of a frame according to an embodiment of the application;

[0070] [ Figure 3 ] Figure 3 is a diagram of an information exchange scheme between the destination and the nodes (i.e. sources and relays) according to an embodiment of the application. DETAILED DESCRIPTION

[0071] A channel use is the smallest time-frequency resource granularity defined by the system that allows the transmission of a modulation symbol. The number of channel uses is related to the available frequency band and the transmission duration.

[0072] In Figure 1 OMA MRC systems are illustrated. Such a system comprises M sources belonging to a source set L relays belonging to a relay set and one destination d.

[0073] Each source in the set communicates with a single destination with the assistance of other sources (user cooperation) and relays that cooperate.

[0074] To simplify the description, the following assumptions are made below with respect to OMA MRC systems:

[0075] - the sources, the relays are equipped with a single transmit antenna;

[0076] - the sources, the relays and the destination are equipped with a single receive antenna;

[0077] - the sources, the relays and the destination are perfectly synchronized;

[0078] - the sources are statistically independent (no correlation between them);

[0079] - all nodes transmit with the same power;

[0080] - a CRC code is used, assuming that it is included in the K i information bits of each source i in order to determine whether the message is decoded correctly,

[0081] - the links between the various nodes experience additive noise and fading. When a frame is transmitted over a maximum duration of 1 + T max slots, the fading gain is fixed but can change independently from one frame to another. T max ≥ 1 is a parameter of the system;

[0082] - the instantaneous quality of the direct reception channel / link (CSIR "Channel State Information at the Receiver") is available at the destination, the sources and the relays;

[0083] - the feedback is error-free (control signals are error-free).

[0084] The nodes include relays and sources that can act as relays when they do not emit their own message.

[0085] Nodes (i.e., M sources and L repeaters) access the transmission channel according to a frequency orthogonal multiple access scheme and operate in full-duplex mode, which allows them to listen to the transmissions of other nodes without any interference.

[0086] The channel is divided into B sub-bands, assuming the number of sub-bands is greater than or equal to the number of sources: B ≥ M. Each sub-band associated with a time slot determines F channel uses (F resource elements).

[0087] When transmission is performed using OFDM modulation, the subband may include, for example, as many subcarriers as OFDM symbols.

[0088] Assume that for each transmission time slot, the number of channel uses N is the same: N = B × F.

[0089] The transmission period lasts for 1+T 已用 There are 1 time slot, of which T 已用 ≤T max and T max This is the maximum number of time slots. In each time slot, depending on the first partition, either no subband is assigned to a node, or one or more subbands are assigned to a node.

[0090] During the first time slot (first phase), assuming B≥M, all sources transmit on one or more sub-bands allocated to each source.

[0091] During the subsequent "retransmission slot" (second phase), only nodes selected from the source and repeaters retransmit, and their retransmissions occur on one or more subbands allocated to them according to the partitions determined for each current slot. Therefore, partitions can differ across all transmission slots, including the first transmission slot.

[0092] The selection of nodes and the allocation of subbands are carried out by a scheduler, which is typically hosted by the destination.

[0093] Use the following notation:

[0094] • If i ≤ M, then the selected node i is the source i, denoted as s. i If i ∈ {1, ..., M}, otherwise i > M, and the selected node is a repeater iM, denoted as r. i-M , i∈{M+1,...,M+L};

[0095] · This is the dimension vector B of the node selected for transmission slot t, whether this is during the first phase or the second phase. Vector a t The i-th element a t，iSpecify the i-th subband and the selected nodes that are active in subband i during time slot t, i∈{1,...,B}. The order in the vector corresponds to the order of the subbands;

[0096] ·n t ∈{0, ..., B} M+L This is a dimension vector M+L representing the number of subbands allocated to each node (i.e., source or repeater) for a transmission time slot (slot t). This number of subbands varies between 0 (node ​​is inactive) and B (node ​​occupies all subbands), regardless of whether this occurs during the first or second phase. Vector n t The i-th element n t，i This represents the number of subbands allocated to node i in transmission time slot t, where i ∈ {1, ..., M+L}. It forms a vector n. t The sum of the elements equals B, which is the number of subbands;

[0097] ·h a，b It is the attenuation gain (fading) of the channel between node a (source or repeater) and node b (source, repeater, or destination), which follows a path with zero mean and variance γ. a，b The symmetric cyclic complex Gaussian distribution has independent gains;

[0098] ·T 已用 This is the minimum number of retransmission slots (i.e., during the second phase), which results in zero failures for all sources (zero individual cutoff events for each source):

[0099]

[0100] A separate cutoff event for source s after retransmission time slot t (t round). Depends on the vector a used to select the node t The vector n used for sub-band allocation t and the source set decoded at the end of the previous time slot t-1 This individual cutoff event also depends on the understanding of the implementation of the channel on the direct link. dir (Channel gain) and The set of specified selection vectors (and therefore, the selected nodes). and the set of assignment vectors And the associated decoded source set of these vectors determined for the time slot (round) l preceding time slot t. l∈{1,...,t-1} and the source set decoded by the destination It should be noted that a0 is the selection vector of the source node to be transmitted during the transmission phase, and n0 is the allocation vector of the subband assigned to each source during the transmission phase. is the set of sources decoded by the destination upon completion of the first phase.

[0101] is the common cut event for the subset of sources at time slot t (t round) is the event where at least one source in the subset is not correctly decoded by the destination at the end of time slot t. Subsequently, for simplicity of notation, the dependence on h dir and on is omitted. denotes the set of sources not successfully decoded by the destination at the end of time slot t (t round).

[0102] From an analytical point of view, the common cut event for the subset of sources occurs, i.e. is satisfied, if the vector of rates of these sources is not included in the corresponding capacity region MAC.

[0103] Therefore, for a given subset of sources for a candidate vector a t

[0104] and the corresponding sub-band allocation vector n t the event can be expressed in the following form:

[0105]

[0106] where, denotes the non-compliance with the inequality MAC associated with the sum rate of the sources contained in :

[0107]

[0108] where:

[0109] • l is the time slot index (round) of the second phase, by convention l = 0 corresponds to the end of the first phase (transmission phase), l e {1,.., T 已用};

[0110] • s is the index corresponding to a source node, s e {1,..., M};

[0111] • i is the index corresponding to any node (source and relay), i e {1,..., M+L};

[0112] • n l，i is the number of sub-bands allocated to node i at time slot l (round), l e {1,..., T 已用};

[0113] • n O，sis the number of subbands assigned to source s e {1,...,M} by the destination in the first phase;

[0114] ·

[0115] wherein, denotes one if the intersection between the set of sources decoded correctly by node i at time slot l-1 and the set is not empty, and the intersection between the set of sources decoded correctly by node i at time slot l-1 and the set of interfering sources is empty, then the set of interfering sources equals one;

[0116] denotes the "and" logic;

[0117] • [P] denotes the Iverson bracket, i.e. it yields the value 1 if event P is fulfilled, and 0 otherwise;

[0118] · is the mutual fading information block of source s at destination d for subband n 已用 assigned to node i at time slot l e {l,...,T l，i}:

[0119]

[0120] wherein, is the mutual information between node a l，f and destination d for subband f at time slot (round) l e {l,...,T 已用}. The mutual information depends on the power transmitted on the subband of the channel, i.e. the l，f wherein P T is the total power of the node. If node i is not selected at time slot l, the mutual information block is zero;

[0121] · is the mutual fading information block of source s at destination d for the time slot corresponding to the transmitting node (first phase) for given a0and n0;

[0122] • R s = K s / (n 0，s F), s = 1,...,M is the rate used during the first phase, wherein K s is the number of useful information bits transmitted on n 0，s F channel uses.

[0123] Subsequently, the cut event for a given source s is defined in the following form: ​

[0124]

[0125] By definition, this is the intersection of all common cut events corresponding to the set of sources including source s Source s is cut if and only if no set of sources including source s can be associated with a perfect decoding (i.e., ). This cut event becomes:

[0126]

[0127] This cut event indicates whether the source is perfectly decoded or cut This method allows to predict the outcome of the implementation of the parity check (CRC check) without the need to simulate the whole transmission (modulation coding) and reception (detection / demodulation, decoding) chain. In this way, an abstract concept of the physical layer is defined. By introducing a weighting parameter of the mutual information of the link and / or of the SNR, some adjustments (referred to as calibrations in the context of the abstract concept of the physical layer) obtained by simulation can be performed for a given coding scheme.

[0128] Before the two transmission phases of the transmission method, there can be an initial phase of determination of the initial rate. In the case of slow fading, this phase can occur once every few hundred frames (i.e. each time the quality statistics of the channel / link change), which is referred to as slow link adaptation. Alternatively, this phase can occur more frequently and at most at each period, which is referred to as fast link adaptation. Whether the link adaptation is fast or slow, the rate and the allocation of subbands for each source are known before the transmission starts.

[0129] By exploiting reference signals (pilot symbols of the 3GPP LTE / NR DMRS type, reference signals of the 3GPP LTE / NR SRS type, etc.), the destination can determine the gain of the direct link (CSI “Channel State Information”): i.e. source link to destination and relay to destination. Therefore, for the direct link within the context of slow adaptation, the destination can derive the average from it.

[0130] By contrast, the gain of the links between sources, between relays and between sources and relays is not known to the destination. Only the sources and the relays can evaluate the metrics of these links by exploiting reference signals in a similar way as for the direct link.

[0131] Considering that, in the context of slow adaptation, the statistics of the channel are assumed constant between two initialization phases, the transmission metrics from sources and relays to the destination can only occur at the same rate as the initialization phases. The statistics of the channel of each link are assumed to follow a centered circular complex Gaussian distribution and are independent between links.

[0132] In the context of fast adaptation, the optimization of the spectral efficiency can be based on the knowledge of all links or some of them. One possible, but very cumbersome in terms of control, solution is that sources and relays feedback to the destination the link coefficients (quantized per subband) they can evaluate.

[0133] During an initial link adaptation phase before the transmission of one or more frames, the destination transmits a representative value (index, MCS, rate, etc.) of an initial rate for each source s. Each initial rate explicitly determines an initial modulation and coding scheme (MCS), or conversely, each initial MCS determines an initial rate. The initial rates are fed back via a control channel of very limited rate.

[0134] These initial rates are determined by the destination in order to maximize a quality of service metric (e.g. average spectral efficiency).

[0135] In the case of slow link adaptation, according to one embodiment, the quality of service metric is the average spectral efficiency, which is expressed in the form:

[0136]

[0137] where:

[0138] · is a variable representing the initial rate allocated to source i, i e {1,...,M};

[0139] · K i is the amount of data transmitted by source i on n 0，i x F channel uses;

[0140] · T 已用 is the number of time slots for cooperative or non-cooperative retransmission;

[0141] · is the average number of time slots for retransmission, whether cooperative or non-cooperative;

[0142] · BLER i is the average block error rate of source i.

[0143] In the case of slow link adaptation, the rate and subband allocation of each source remains constant for several hundreds of message transmissions from the source, which allows to average the block error rate (BLER) of source i over the channel statistics known at the destination (CDI "Channel Distribution Information"). Source i acquires n O，i x F resource elements in order to transmit data K i at rate R i .

[0144] In the case of fast link adaptation, according to one embodiment, the quality of service metric is an average utility function subject to individual cutoffs of the sources defined per transmitted message, and the rate and subband allocation can change from one message to the next:

[0145]

[0146] where:

[0147] · is the individual cutoff event of source i at retransmission slot t, which equals one in case of failure or zero in case of success (correctly decoded source), is the individual cutoff event at the end of the transmission phase (first phase of the slot);

[0148] ·T 已用 is the number of slots for cooperative or non-cooperative retransmission (the second phase can assume that if then the value is 0);

[0149] · is a variable representing the initial rate allocated to source i, i e {1,...,M}.

[0150] Reference is made to Figure 2 Figure 1 describing one embodiment of a transmission method according to the application, which illustrates a transmission period of a frame within the context of an OMA MRC system having three sources (i = {1,2,3}), two relays (i = {4,5}), one destination and a transmission channel having a specific bandwidth. The band is divided into B = 5 subbands and each subband associated with a slot determines F = 5 channel uses, that is to say, N = 25.

[0151] During the first phase of the slot, each source i = {1,2,3} emits its codeword. According to this example, the number of subbands allocated to the sources is different between the sources. Thus, subbands f1, f2 and f5 are allocated to source 1, subband f3 is allocated to source 2 and subband f4 is allocated to source 3. The selection vector is therefore a0= [s1, s1, s2, s3, s1] T = [1, 1, 2, 3, 1] T. Thus, the vector for allocating subbands to each node is n0= [3, 1, 1, 0, 0] T .

[0152] During the second phase, called the retransmission phase, and for the first time slot, only sources 2, 3 and repeater 2 are selected, and subband fl is allocated to source 3, subbands f2, f3 and f4 are allocated to repeater 5, and subband f5 is allocated to source 2. Thus, the selection vector is ai = [s3, r2, r2, r2, s2] T = [3, 5, 5, 5, 2] T . Thus, the vector for allocating subbands to each node is ni = [0, 1, 1, 0, 3] T .

[0153] During the second phase, called the retransmission phase, and for the second time slot, only sources 1 and repeater 4 are selected, and subbands fl, f2 and f5 are allocated to repeater 4, subbands f3 and f4 are allocated to source 1. Thus, the selection vector is a2= [r1, r1, s1, s1, r1] T = [4, 4, 1, 1, 4] T . Thus, the vector for allocating subbands to each node is n2= [2, 0, 0, 3, 0] T .

[0154] Figure 3 An embodiment of the exchange scheme between the nodes and the destination is illustrated in Fig. 3.

[0155] Each source transmits its frame data to the destination with the help of the other sources and repeaters. The frame occupies a time slot respectively during the transmission of the M messages of the M sources. The transmission of the frame, which defines a transmission period, occurs within 1 + T 已用 time slots: 1 time slot for phase 1, where for each source i, the channel usage capacity is n O，i , T 已用 time slots for phase 2, where for each source i, the channel usage capacity is n t，i .

[0156] During the first phase, each source transmits a message u s including K s information bits after encoding, s where K s is the number of information bits, and u s is a double-element Galois field. The message u 已用 includes a code of CRC type which allows to check the integrity of the message u 已用 .are encoded according to the initial MCS. Considering that the initial MCS can be different between sources, the encoded message length can also be different between sources. The encoding uses an incremental redundancy code. The obtained codeword is segmented into redundancy blocks. The incremental redundancy code can be of systematic type, then the information bits are included in the first block. Whether the incremental redundancy code is of systematic type or not, it makes the first block decodable independently of the other blocks. The incremental redundancy code can be produced for example by a rate-compatible punctured linear code or a rateless code modified to operate with finite length: raptor code (RC), rate-compatible punctured turbo code (RCPTC), rate-compatible punctured convolutional code (RCPCC), rate-compatible low-density parity-check code (RCLDPC).

[0157] Thus, during the first phase, the M sources simultaneously transmit their messages on the allocated subbands using modulation and coding schemes determined according to the initial rate values, respectively, according to the vector a0during the transmission time slot.

[0158] Each transmitted message (correctly decoded message) of a source is assimilated with the corresponding source in order to be labelled.

[0159] Whatever it is, during the first phase or during the second phase, when a node (in particular a source) transmits, the destination and the other nodes listen. Considering that each node is allocated one or more subbands different between nodes for transmission, each full-duplex node can transmit and listen to all the other nodes simultaneously.

[0160] The destination, the sources and the relays try to decode the messages received at the end of the time slot. The success of the decoding at each node is determined using the CRC. Thus, the destination and the nodes determine their correctly decoded source set.

[0161] During the second phase, at time slot (round) t, the destination d transmits its correctly decoded source set at the end of the previous time slot t 已用 using for example a feedback broadcast control channel t = {1,..., T The feedback can consist of a vector of M bits.

[0162] If the decoding of all the sources by the destination is correct, then In this case, the current period is stopped and a new period can start. The period for transmitting a new frame starts with the deletion of the memories of the relays and of the destination and the transmission of new messages by the sources. The number of time slots (rounds) T 已用 = {1,..., T max used during the second phase depends on the success of the decoding at the destination.

[0163] Nodes, sources and relays will collect with their correctly decoded source set.

[0164] If the node set includes at least one source not included in the destination set , the node informs the destination accordingly by using a dedicated control channel of e.g. unicast type. The information transmitted by the node can consist of its correctly decoded source set, or as shown in Figure 3 , of e.g. a bit set to one.

[0165] During this second phase, the destination follows a specific strategy in order to determine one or more selected nodes that transmit at time slot (round) t.

[0166] The destination informs the nodes of this selection by transmitting a vector a t using e.g. a feedback broadcast control channel.

[0167] Each node receiving the vector a t can determine whether the node is selected and on which subband the node will transmit.

[0168] During this second phase, and for at least one of the T 已用 retransmission time slots, at least one selected node, source or relay generates a cooperative retransmission. Outside this at least one time slot, the retransmission can be cooperative or non-cooperative.

[0169] The nodes selected for retransmission transmit a parity determined based on the messages of their correctly decoded source set after multi-user encoding of the words or partial words they have correctly decoded. Other nodes and the destination can improve their own decoding by exploiting the transmission of the selected nodes and can thus update their correctly decoded source set.

[0170] The destination thus uses a feedback channel to control the transmission of the nodes. This allows to increase the spectral efficiency and reliability by increasing the probability of the destination to decode all sources.

[0171] Selection strategy

[0172] According to a first strategy, the destination selects from the set of various subbands Zt that can be allocated to nodes that can assist at time slot t, those subbands that maximize the sum of mutual information. This strategy only requires knowledge of the nodes that can assist, and is compatible with the mode of transmission of the nodes in the form of bits.

[0173] The selection criterion can be expressed in the form:

[0174]

[0175] where Z t is the set of possible vectors a t corresponding to the selection of a node that can assist the destination at time slot (round) t.

[0176] Based on the correctly decoded source set received from the nodes and according to a second strategy, the destination selects the node that allows obtaining as many new correctly decoded sources at the destination at the end of the current time slot t, i.e. the node that maximizes the cardinality of the set of correctly decoded sources at the destination at the end of the current time slot t.

[0177] According to this strategy, the method examines all possible values of the vector a t and retains the value that results in the maximum number of new decoded sources. Thus, the method does not take into account the nodes that cannot assist a source that has not yet been decoded, since the objective of the method is the maximum number of new decoded sources, i.e. only the nodes i that satisfy the following condition are considered:

[0178] This strategy also requires knowledge of the correctly decoded source set of all previously selected nodes.

[0179] In the case where several vectors a t can result in the same maximum number of new decoded sources, the method selects the vector that maximizes the sum of mutual information In fact, at time slot t, this is the only element that can be maximized to maximize the straight part of the individual cutoff event and the common cutoff event. The presence in the expression of the common cutoff event represents the fact that only the nodes that can assist, i.e. the nodes that have already decoded at least one source that has not yet been decoded by the destination, can be selected. The selection criterion can then be expressed in the form:

[0180]

[0181] where, is the set of candidate nodes that maximizes the destination set at the end of time slot (round) t.

[0182] It should be noted that for t = 0, the only candidate nodes of the first phase are the sources, whose decoded set corresponds to itself, and the relay nodes have an empty decoded set.​

Claims

1. A method for transmitting a frame message using an orthogonal multiple access to the transmission channel called OMAMRC, this method being intended for a telecommunication system having M sources , L relays and one destination , , , characterized in that The transmission is of frequency division multiplexing (FDM) type and is made on a band divided into B mutually orthogonal subbands, and the method is such that it comprises: - the simultaneous transmission of the M sources during a certain time slot, each source being allocated at least one subband; and - at least one cooperative retransmission by at least one relay node selected by the destination from the M sources and the L relays during a certain time slot, the destination allocating at least one subband to each selected relay node, the destination selecting the relay nodes that allow the most new correctly decoded sources at the destination at the end of the cooperative retransmission by knowing the sources correctly decoded by the relay nodes.

2. The method of claim 1, wherein, The selection by the destination is such that a relay node that decodes a set of sources at time slot t can only perform a cooperation at time slot t+1 for a single source of its set.

3. The method of claim 1, wherein: - the destination broadcasts to the relay nodes its set of correctly decoded sources from the received sources during the transmission time slot; - the relay nodes that have correctly decoded a source that was not correctly decoded by the destination inform the destination of this fact; - the destination broadcasts a vector to the relay nodes the relay nodes comprise relay nodes selected for the subbands for cooperative retransmission or non-cooperative retransmission during the next transmission time slot.

4. The method of claim 3, wherein, The relay nodes inform the destination by transmitting their set of correctly decoded sources.

5. The method of claim 1, wherein, The destination selects the relay nodes such that the sum of mutual information between the nodes that can assist with the subbands allocated to them and the destination is maximized.

6. The method of any one of the preceding claims, having an initial phase with a determined initial rate and such that the initial rate allocated to the sources is determined to maximize a metric expressed in the form of an average utility function subject to an average individual BLER for each source: , wherein: • is a variable representing the initial rate allocated to this source , where, represents the number of subbands allocated to this source i at time slot 0, represents the number of channel uses for each subband associated with the time slot; • is the amount of data transmitted by the source i on the secondary channel usage; • is the number of slots for cooperative or non-cooperative retransmission; • is the average number of slots for cooperative or non-cooperative retransmission; • is the block error rate of the source i.

7. The method of any one of claims 1 to 5, having an initial phase with a determined initial rate per frame and such that the initial rate allocated to the sources is determined to maximize a metric expressed in the form of an average utility function subject to individual cutoffs for the sources: wherein: • is the individual outage probability of the source i at the cooperative or non-cooperative retransmission time slot t; • is the number of cooperative or non-cooperative retransmissions; • is a variable representing the initial rate assigned to this source ;​ • is the number of subbands allocated to the source i in the time slot is the number of subbands allocated to the source i in the time slot ; • is the number of channel uses.

8. A system comprising one source , one repeater and one destination , , adapted to carry out the method of any one of claims 1 to 7.

Images