Distributed multi-node relay communication system wireless space-frequency channel sensing method based on error feedback

Abstract

The invention relates to the technical field of wireless communication, in particular to a distributed multi-node relay communication system wireless space-frequency channel sensing method based on error feedback, which comprises the following steps: a line transmitter configures the initial state of each distributed wireless relay node through a control signaling channel; the wireless transmitter frames the training sequence and the communication signal and sends the framed training sequence and communication signal to an intermediate node in a coverage range; the distributed relay node classifies, weights and forwards the received training sequence and the communication signal; the method has the beneficial effects that in the iteration process, only the initial state information of each node needs to be synchronized, and only the single normalized channel matching error feedback of a receiving end is needed for sensing and weight updating of the wireless space-frequency channel, so that the distributed relay nodes do not need to interact and receive results mutually; and each node can synchronously perceive through iteration to obtain a space-frequency weighting coefficient of the whole system, and further perceive a wireless space-frequency channel of the system.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technologies, and particularly to a method for wireless spatio-frequency channel sensing of a distributed multi-node relay communication system based on error feedback. Background Art

[0002] In order to prevent the signals of each relay link from being aliased and difficult to distinguish at the receiver, the weight adjustment of a traditional wireless multi-node relay system usually requires channel estimation for each relay link between the transmitter, relay nodes, and receiver one by one, and then the updated weights are fed back to each relay node through a signaling control link. In the case of limited spatial coverage, there is also a method of constructing a suitable weight codebook to simplify the process of updating relay weights between the transceiver.

[0003] However, when implementing these methods, not only a large amount of resources such as time, frequency, and spatial domain are required, but also additional complex designs are required for the transceiver and each relay node in terms of communication frame structure, feedback link planning, and hardware resource implementation. At the same time, the channel response estimation and relay weight configuration of each relay link will increase significantly with the increase in the number of relay nodes and the number of antennas configured for each relay, and it is difficult to ensure the real-time performance of wireless channel sensing or target node position sensing in a dense dynamic scenario. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for wireless spatio-frequency channel sensing of a distributed multi-node relay communication system based on error feedback to solve the problems raised in the above background art.

[0005] To achieve the above purpose, the present invention provides the following technical solution: A method for wireless spatio-frequency channel sensing of a distributed multi-node relay communication system based on error feedback, the method comprising the following steps:

[0006] Step 1, the wireless transmitter configures the initial states of each distributed wireless relay node through a control signaling channel;

[0007] Step 2, the wireless transmitter frames the training sequence and communication signals and sends them to the intermediate nodes within the coverage range;

[0008] Step 3, the distributed relay nodes classify, weight, and forward the received training sequence and communication signals;

[0009] Step 4, the wireless receiver obtains the normalized error through channel estimation and broadcasts the feedback;

[0010] Step 5, the distributed wireless relay nodes iteratively update the spatio-frequency domain weights using the normalized feedback error;

[0011] Step 6, the communication system uses the weights fed back by the distributed nodes to achieve spatio-frequency channel and user position sensing.

[0012] Preferably, step 1 specifically includes:

[0013] For a distributed multi-node wireless relay communication system composed of M relay node devices, the degrees of freedom of its wireless link are usually composed of the number of antennas N of each node, the number of subcarriers P supported by the relay nodes, and the number of relay nodes M. Before operation, the wireless transmitter configures the initial working state of each distributed wireless relay node through the control signaling channel, so that the iteration count k of each current node is 0; without loss of generality, when the iteration count value k = 0, the space-frequency beam weights w for coherent combining of all wireless relay nodes 0 = 1 NPM×1 ; while when k ≠ 0, the space-frequency weights w for coherent combining of each distributed wireless relay node k,m (m = 1, …, M) and the weight relationship of the entire distributed system can be expressed as:

[0014]

[0015] where, (·) T represents the transpose operation of each element in a scalar, vector, or matrix; the space-frequency weight w of the m-th node in the k-th iteration k,m is then composed of the P subcarrier weights corresponding to the N antennas, as shown in the following formula:

[0016]

[0017] Preferably, step 2 specifically includes:

[0018] After the initialization configuration of the wireless relay nodes is completed, the distributed wireless relay communication system led by the wireless transmitter starts to enter the communication state of space-frequency channel sensing and transmitting useful information to the receiver. During this process, the wireless transmitter frames, which includes two consecutive training sequences c pre and c pro for synchronization and weighting of intermediate nodes within the coverage range, and several communication symbol blocks s q (q = 1, …, Q) for transmitting information.

[0019] Preferably, step 3 specifically includes:

[0020] The wireless signal from the transmitter is transmitted through the space-frequency channel and is jointly represented in the spatial and frequency domains when it reaches the receiver of the m-th relay node as:

[0021]

[0022] In the above formula, n m,pre,k 、n m,pro,k and n m,q,kThey respectively represent the noise terms when receiving two training sequences and communication signals, and the subscript k indicates the number of receptions in the current iteration. represents the Kronecker product between matrices or vectors, and diag(·) represents diagonalizing a vector or taking the diagonal elements of a square matrix and vectorizing them.

[0023] Each distributed wireless relay node weights the currently received signal according to the current iteration number k in accordance with for c pre and s q for the corresponding parts, and weights the c pro part with the null space weight v k There is the following relationship between this null space weight and the coherent combining weight w k :

[0024]

[0025] where, (·) H represents the conjugate transpose operation on each element in a scalar, vector or matrix, and v k is jointly determined and generated in advance by each wireless relay node according to the current iteration reception number k, the number of relay nodes M, the number of subcarriers P, and the number of relay node antennas N. The space-frequency pre-coding codebook T = (t 1 ,, t NPM ), where t mod(k,NPM) is the mod(k, NPM)-th column of the codebook matrix T, and ||t mod(k,NPM) || = 1. This space-frequency codebook is pre-generated according to the rules and is known to all wireless relay nodes in the system. Thus, v k is obtained synchronously and is in the null space of w k , that is, it satisfies

[0026] Preferably, referring to the corresponding relationship between w k and w k,m , the column vector v k used for nulling suppression is also split as follows:

[0027]

[0028] After being weighted by each distributed node in the space-frequency domain, the training sequence and communication signal are timely forwarded. Then, the weighted signals can be expressed as:

[0029]

[0030] where, (·) * represents the conjugate operation on each element in a scalar, vector or matrix.

[0031] Preferably, step 4 specifically includes:

[0032] The signals weighted and forwarded by each distributed relay node pass through various wireless channels between the relay node and the receiver and then are superimposed on each other at the receiver antenna. Each part finally forms the following form on the p-th subcarrier:

[0033]

[0034] where

[0035]

[0036] are all simplified notations of "starting from p elements and taking values at intervals of P elements from the middle of the NP-dimensional column vector". Similar notations also apply to the noise terms n m,pre,k 、n m,pro,k and n m,q,k in the above formula. After taking values at this interval, the dimension of the column vector is the same as the number of antennas N of the relay node. Rearrange the signal form received on the p-th subcarrier above and further simplify it to:

[0037]

[0038] Here, ⊙ is the Hadamard product between vectors or matrices of the same dimension. It can be seen from this that after being forwarded by the relay communication system, the communication link channel effects of the relay node and the transceivers at both ends of the link are multiplicative. This indicates that to ensure maximum SNR reception, the system processing weights can comprehensively consider the effects of both ends of the relay link transceiver during design without the need for segmented processing.

[0039] Preferably, two estimates of "coherent combining channel sum" and "nulling channel sum" are obtained:

[0040]

[0041] It can be seen from the above formula that h k,pre,∑ contains the complex amplitude effects of all relay space-frequency channels. Therefore, on this basis, a normalized feedback error is constructed:

[0042]

[0043] Since the above formula is taken from the channel estimation results of two consecutive training sequences, in the case of good synchronization at the receiving end, the influence of the initial carrier phase difference between the transmitter and the receiver is eliminated by dividing each other, and only the role of the relay transceiver channel is reflected therein;

[0044] After the wireless receiver demodulates and calculates the feedback error, the feedback error e k . is fed back to all relay nodes through the signaling control link or other wired or wireless communication channels

[0045] Preferably, step 5 specifically includes:

[0046] After each wireless relay node in the system receives the normalized error feedback e from the wireless receiver k , it starts to iteratively update the previous space-frequency coherent combining weights and nulling weights, and the process is shown as follows:

[0047]

[0048] where μ is the step coefficient for the system to iteratively sense the wireless space-frequency channel, usually less than 1, and here μ = 0.5; v k+1 is the mod(k + 1, NPM)-th column of the codebook matrix T, and ||t mod(k+1,NPM) || = 1; repeat steps 2 - 5 according to the above formula to obtain gradually convergent and stable coherent combining and nulling space-frequency beam weights, and each relay node only needs to take the elements corresponding to its own node number in w k and v k for weighted forwarding of the received signal;

[0049] For w k and v k are presented in the form of space-frequency weights of all relay nodes in the whole system. Through the configuration of the initial state in step 1 and the constraint of each node number, each relay node can obtain the minimum mean square error weighted coefficient of the whole system for the wireless space-frequency channel without interaction.

[0050] Preferably, step 6 specifically includes:

[0051] As the number of iterations of the transceiver communication and feedback error weights increases, the coherent combining weights of the distributed multi-node relay system will gradually converge in the direction of the minimum mean square error to be in the same direction as the space-frequency channel h (t) ⊙h (r) . The transceiver channels of the system relay link can be respectively expressed by the following formula:

[0052]

[0053] When the iteration error has converged to a small value, any distributed relay node sends the currently converged coherent combining weight w k to the wireless transmitter or receiver, so that the transceiver nodes obtain the sensing estimate of the current wireless space-frequency transmission channel h (t) ⊙h (r) .

[0054] Preferably, when the transmitter knows the positions and postures of its own transmitting antennas and the antenna arrays of each distributed node, h (r)Known to the transmitter, the angle and time delay of the target receiving device relative to the m-th distributed node are obtained by the following formula:

[0055]

[0056] where a m (θ) and g(τ) are respectively the spatial domain steering vector of the m-th distributed relay node and the frequency domain channel response caused by the time delay τ. Based on the positions of each relay node, the coordinate position (x r , y r , z r ) of the target receiver node is sensed through joint DOA-TDOA estimation.

[0057] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0058] The wireless space-frequency channel sensing method of the distributed multi-node relay communication system based on error feedback proposed by the present invention only requires the synchronization of the initial state information of each node during the iteration process. The sensing and weight update of the wireless space-frequency channel only require the single normalized channel matching error feedback at the receiving end. It not only does not require the distributed relay nodes to exchange their respective receiving results with each other, but also enables each node to synchronously sense the space-frequency weighting coefficients of the entire system through iteration. At the same time, the present invention occupies less link resources and control information, and while making full use of the ultra-large space-frequency observation aperture of the distributed relay communication system, it can realize the continuous sensing of the downlink channel or the target position in a dynamic scenario. In addition, the present invention only requires the single error feedback at the receiving end for the wireless channel sensing and weight update, without the independent measurement of each relay link between each transceiver node, which can ensure the synchronization of the update of the space-frequency weights of the distributed nodes. Description of the Drawings

[0059] Figure 1 It is a block diagram of the distributed multi-node wireless relay communication system with a control signaling channel according to the present invention;

[0060] Figure 2 It is a schematic diagram of the wireless space-frequency channel sensing process of the distributed multi-node relay communication system based on error feedback according to the present invention;

[0061] Figure 3 It is a schematic diagram of the training sequence and communication signal frame structure to be transmitted according to the present invention. Detailed Embodiment

[0062] In order to clearly and completely describe the objectives, technical solutions of the present invention, and make the advantages more clearly understood, the following further details the embodiments of the present invention with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are part of the embodiments of the present invention, rather than all of the embodiments, and are only used to explain the embodiments of the present invention, not to limit the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0063] Please refer to Figures 1 to 3 , the present invention provides a technical solution: a wireless space-frequency channel sensing method for a distributed multi-node relay communication system based on error feedback. The method includes the following steps:

[0064] Step 1: The wireless transmitter configures the initial states of each distributed wireless relay node through the control signaling channel

[0065] For a distributed multi-node wireless relay communication system composed of M relay node devices, the degrees of freedom of its wireless link are usually composed of the number of antennas N of each node, the number of subcarriers P supported by the relay nodes, and the number of relay nodes M. That is to say, the number of transceiver communication links that the wireless relay system can support simultaneously in space is not simply limited by the number of antennas of each relay node. In the present invention, in order to reduce or even eliminate a large amount of time-consuming information interaction between distributed nodes, here the wireless transmitter configures the initial working states of each distributed wireless relay node through the control signaling channel before work, so that the iteration count k of the current nodes is 0. Without loss of generality, when the iteration count value k = 0, the space-frequency beam weights w 0 = 1 NPM×1 ; and when k ≠ 0, the space-frequency weights w k,m (m = 1,..., M) of each distributed wireless relay node for coherent combination and the weight relationship of the entire distributed system can be expressed as

[0066]

[0067] Among them, (·) T represents the transpose operation of each element in a scalar, vector, or matrix; the space-frequency weight w k,m of the m-th node in the k-th iteration is composed of the weights of P subcarriers corresponding to N antennas, as shown in the following formula:

[0068]

[0069] Step 2: The wireless transmitter frames the training sequence and communication signals and sends them to the intermediate nodes within the coverage range

[0070] After the initialization configuration of the wireless relay node is completed, the distributed wireless relay communication system led by the wireless transmitter starts to enter the communication state of sensing the empty frequency channel and transmitting useful information to the receiver. During this process, the wireless transmitter frames according to the frame structure shown in the following figure, which includes two consecutive training sequences c pre and c pro for synchronization and weighting of intermediate nodes within the coverage area, and several communication symbol blocks s q (q = 1,..., Q) for transmitting information.

[0071] Step 3: Classify, weight, and forward the received training sequence and communication signal by the distributed relay node

[0072] The wireless signal from the transmitter passes through the empty frequency channel and can be jointly represented in the spatial and frequency domains when it arrives at the receiver of the m-th relay node as:

[0073]

[0074] In the above formula, n m,pre,k , n m,pro,k and n m,q,k represent the noise terms when receiving the two training sequences and the communication signal respectively, and the subscript k represents the current iteration reception times. denotes the Kronecker product between matrices or vectors, and diag(·) represents diagonalizing a vector or taking the diagonal elements of a square matrix and vectorizing them.

[0075] After that, Figure 1 each distributed wireless relay node in the system weights the currently received signal according to the current iteration times k for the corresponding parts of c and s pre and s q , while weighting the c pro part with the null space weight v k . There is the following relationship between this null space weight and the coherent combining weight w k :

[0076]

[0077] where (·) H represents the conjugate transpose operation on each element of a scalar, vector, or matrix. v k is pre-determined and generated jointly by each wireless relay node according to the current iteration reception times k, the number of relay nodes M, the number of sub-carriers P, and the number of relay node antennas N, and the empty frequency pre-coding codebook T = (t 1 ,, t NPM ), where t mod(k,NPM)is the mod(k,NPM)-th column of the codebook matrix T, and ||t mod(k,NPM) || = 1. This space-frequency codebook is pre-generated according to the rules and is known to all wireless relay nodes in the system. Thus, v can be obtained synchronously k in the null space of w k , that is, satisfying

[0078] For the convenience of subsequent discussion, referring to the corresponding relationship between w k and w k,m , the column vector v k used for nulling suppression is also split as follows:

[0079]

[0080] After being weighted by each distributed node in the space-frequency domain, the training sequence and the communication signal are forwarded in a timely manner, and the weighted signals can be expressed as:

[0081]

[0082] where (·) * represents the conjugate operation of each element in a scalar, vector, or matrix.

[0083] Step 4: The wireless receiver obtains the normalized error through channel estimation and broadcasts the feedback

[0084] Figure 1 The signals weighted and forwarded by each distributed relay node shown pass through various wireless channels between the receiver and then are superimposed on each other at the receiver antenna, and their respective parts finally form the following expression on the p-th subcarrier:

[0085]

[0086] where

[0087]

[0088] are all simplified notations for "starting from the p-th element and taking values at intervals of P elements from the middle of the NP-dimensional column vector". Similar notations also apply to the noise terms n m,pre,k , n m,pro,k and n m,q,k in the above formula. The dimension of the column vector after taking values at this interval is the same as the number of antennas N of the relay node. It should be noted that in the above expression, considering the cascade system noise coefficient, the influence of the receiver noise is temporarily ignored, and only the input noise of the relay system is considered. After organizing the signal form received on the p-th subcarrier above, it can be further simplified as:

[0089]

[0090] Here, ⊙ is the Hadamard product between vectors or matrices of the same dimension. It can be seen from this that after being forwarded by the relay communication system as shown in Figure 1 the communication link channel effects of the relay node and the transceivers at both ends of the link are multiplicative. This indicates that in order to ensure maximum SNR reception, the system processing weights can comprehensively consider the effects of both ends of the relay link transceiver during design, rather than processing in segments.

[0091] Since the frequency-domain response of the receiver to the training sequence is known, it is easy to eliminate the influence of the training sequence component from the above formula, and thus obtain two estimates of the "coherently combined channel sum" and the "null-steering channel sum":

[0092]

[0093] It can be seen from the above formula that h k,pre,∑ contains the complex amplitude effects of all relay space-frequency channels. Therefore, a normalized feedback error can be constructed based on this:

[0094]

[0095] Because the above formula is taken from the channel estimation results of two consecutive training sequences, in the case of good synchronization at the receiving end, the influence of the initial carrier phase difference between the transmitter and the receiver can be eliminated by dividing each other, and only the role of the relay transceiver channel is reflected therein.

[0096] After the wireless receiver demodulates and calculates this feedback error, the feedback error e k is fed back to all relay nodes through the signaling control link or other wired or wireless communication channels.

[0097] Step 5: The distributed wireless relay nodes iteratively update the space-frequency domain weights using the normalized feedback error

[0098] After each wireless relay node in the system receives the normalized error feedback e k from the wireless receiver, it begins to iteratively update the previous space-frequency coherent combining weights and null-steering weights, and the process is shown as follows:

[0099]

[0100] Among them, μ is the step coefficient for the system to iteratively sense the wireless space-frequency channel, usually less than 1, and here μ = 0.5; v k+1 is the mod(k + 1, NPM)-th column of the codebook matrix T as described above, and ||t mod(k+1,NPM)|| = 1. Repeat steps 2 - 5 according to the above formula, and gradually convergent and stable coherent combining and nulling suppression spatial - frequency beam weights can be obtained. The weight iterative calculation process only requires vector calculation and does not require matrix inversion or other high - complexity algorithms. When each relay node is in use, only the elements corresponding to its own node number in w k and v k are taken to weight and forward the received signal.

[0101] It should be noted that here, w k and v k are presented in the form of spatial - frequency weights of all relay nodes in the whole system. This means that in the iterative process of the method proposed in the present invention, through the configuration of the initial state in step 1 and the constraint of each node number, each relay node can obtain the minimum mean - square - error weighted coefficient of the whole system for the Figure 1 shown wireless spatial - frequency channel without interaction. This is beneficial to the low complexity and real - time processing of the wireless spatial - frequency channel of the relay system and the wireless perception of the target receiving position.

[0102] Step 6: The communication system uses the distributed node - feedback weights to realize spatial - frequency channel and user - position perception

[0103] As the number of iterations of the transceiver communication and feedback error weights increases, Figure 1 the coherent - combining weights of the shown distributed multi - node relay system will gradually converge in the direction of the minimum mean - square error and be in the same direction as the spatial - frequency channel h (t) ⊙h (r) . Here, the transceiver channels of the system relay link can be respectively expressed by the following formula:

[0104]

[0105] This means that when the iterative error has converged to a relatively small value, as Figure 2 shown, by sending the currently converged coherent - combining weight w k from any distributed relay node to the wireless transmitter or receiver, the transceiver node can obtain the perception estimate of the current wireless spatial - frequency transmission channel h (t) ⊙h (r) .

[0106] Furthermore, when the transmitter knows the positions and postures of its own transmitting antennas and the antenna arrays of each distributed node, h (r) is known to the transmitter, and the angles and time delays of the target receiving device relative to the m - th distributed node can be obtained through the following formula:

[0107]

[0108] where, a m$(θ)$ and $g(τ)$ are respectively the spatial domain steering vector of the $m$-th distributed relay node and the frequency domain channel response caused by the time delay $τ$. Based on the positions of each relay node, the coordinate position $(x r , y r , z r )$ of the target receiver node can be sensed through joint DOA-TDOA estimation.

[0109] Each device of the system of the present invention needs to have the following basic capabilities:

[0110] Wireless transmitter device: This transmitter device is used to realize the wireless transmission of training sequences, wireless communication, and the sending of control signaling within the working frequency band. This device has the functions of digital baseband shaping, generation, framing, up-conversion, and amplification of training sequences and wireless communication signals in terms of wireless communication, and has the ability to control interaction and synchronization with each distributed relay node in a wired or wireless form through a signaling channel in terms of control signaling. At the same time, in order to realize wireless channel estimation and user receiver position sensing, this wireless transmitter device also has the ability to utilize, analyze, and process the weight information from the relay nodes and realize the estimation of the spatial domain and time delay information of the target receiver.

[0111] Distributed relay node device: The distributed node device in this method is located between the wireless transmitter device and the wireless receiver device and forwards the wireless signals from the transmitter to the receiver in a wireless communication form. Each distributed relay node has no less than one transceiver antenna, has the functions of synchronization and space-frequency weighting processing for the training sequences and wireless signals from the transmitter, and has the ability to perform iterative calculation of space-frequency weight using the normalized error information from the receiver to facilitate the coherent combination of the forwarded wireless signals at the receiver antenna. At the same time, this relay device also has the ability to interact with the wireless transmitter and receiver devices in a wired or wireless form to control signaling to realize the transmission of control information, synchronization information, error feedback information, and wireless channel estimation information.

[0112] Wireless receiver device: This receiver device is deployed within the wireless coverage range of the distributed relay system and receives and processes the wireless signals from each distributed relay node. This receiver device has the ability to calculate and generate normalized error feedback by using the space-frequency weight corresponding to the training sequence for coherent combination and nulling suppression of the relay node, and supports communication with each relay node in a wired or wireless manner to broadcast the measurement error feedback to each distributed relay node to facilitate the latter to form iterative updates of the space-frequency weight and space-frequency channel sensing using the error feedback.

[0113] Although embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A wireless space-frequency channel sensing method for a distributed multi-node relay communication system based on error feedback, characterized in that: The method comprises the following steps: Step 1, the wireless transmitter configures the initial state of each distributed wireless relay node through a control signaling channel; Step 2, the wireless transmitter frames the training sequence and the communication signal and sends it to the intermediate node within the coverage area; Step 3, the distributed relay node classifies and weights the received training sequences and communication signals for forwarding; Step 4: The wireless receiver obtains the normalized error through channel estimation and broadcasts feedback; Step 5, the distributed wireless relay node iteratively updates the space-frequency domain weights using the normalized feedback error; Step 6: The communication system uses distributed node feedback weights to achieve space-frequency channel and user location awareness.

2. The method for sensing wireless space-frequency channels in a distributed multi-node relay communication system based on error feedback according to claim 1, characterized in that: Step 1 specifically includes: For a distributed multi-node wireless relay communication system composed of M relay node devices, the degree of freedom of its wireless link is usually composed of three aspects: the number of antennas N of each node, the number of subcarriers P supported by the relay node, and the number of relay nodes M. Before working, the wireless transmitter configures the initial working state of each distributed wireless relay node through the control signaling channel, so that the iteration count k of each current node is 0; without loss of generality, when the iteration count value k = 0, the space-frequency beam weight w0 of all wireless relay nodes for coherent combining is 1 NPM×1 When k≠0, the space-frequency weight w used by each distributed wireless relay node for coherent combining is k,m The weight relationship between (m=1,…,M) and the entire distributed system can be expressed as: in,(·) T represents the transpose operation of each element in a scalar, vector or matrix; the space-frequency weight w of the mth node in the kth iteration k,m It is composed of P subcarrier weights corresponding to N antennas, as shown in the following formula:

3. The method for wireless space-frequency channel sensing in a distributed multi-node relay communication system based on error feedback according to claim 1, characterized in that: Step 2 specifically includes: After the initialization configuration of the wireless relay node is completed, the distributed wireless relay communication system starts to enter the communication state of space-frequency channel perception and transmission of useful information to the receiver, led by the wireless transmitter. During this process, the wireless transmitter performs framing, which includes two consecutive training sequences c pre and c pro Used for synchronization and weighting of intermediate nodes within the coverage area, as well as several communication symbol blocks s q (q=1,…,Q) is used to transmit information.

4. The method for wireless space-frequency channel sensing in a distributed multi-node relay communication system based on error feedback according to claim 1, characterized in that: Step 3 specifically includes: The wireless signal from the transmitter passes through the empty frequency channel The transmission, when arriving at the mth relay node receiver, is represented in the joint spatial and frequency domains as: In the above formula, n m,pre,k 、n m,pro,k and n m,q,k They represent the noise terms when receiving two training sequences and communication signals respectively, and the subscript k represents the number of times the current iteration is received. represents the Kronecker product between matrices or vectors, and diag(·) means to diagonalize the vector or take the diagonal elements of the square matrix and vectorize them; Each distributed wireless relay node processes the currently received signal according to the current iteration number k. C pre and q The corresponding part of is weighted, and c pro Part of the zero-sink space weight v k The null space weight is weighted with the coherent merging weight w k The following relationship exists: in,(·) H represents the conjugate transpose operation of each element in a scalar, vector, or matrix, v k In advance, each wireless relay node pre-codes the space-frequency codebook T = (t1,,t NPM ) jointly determine the generation, where t mod(k,NPM) is the mod(k,NPM)th column of the codebook matrix T, and ||t mod(k,NPM) ||=1, this empty frequency codebook is generated in advance according to the rules and is known to all wireless relay nodes in the system, thus synchronizing to obtain v k In w k In the null space of 5. The method for sensing wireless space-frequency channels in a distributed multi-node relay communication system based on error feedback according to claim 4, characterized in that: Refer to w k and w k,m Here, the column vector v used for zero-falling suppression is k Make the following split: After each distributed node is weighted in the space-frequency domain, the training sequence and the communication signal are forwarded in a timely manner, and the weighted signals can be expressed as: in,(·) * Represents the conjugation operation on the elements of a scalar, vector, or matrix.

6. The method for wireless space-frequency channel sensing in a distributed multi-node relay communication system based on error feedback according to claim 1, characterized in that: Step 4 specifically includes: The signals forwarded by each distributed relay node pass through various wireless channels with the receiver. After that, they are superimposed on each other at the receiver antenna, and their parts finally form the following form on the pth subcarrier: in, Both are simplified notations for "starting from p elements and taking values ​​every P elements from the NP-dimensional column vector". Similar notations also apply to the noise term n in the above formula. m,pre,k 、n m,pro,k and n m,q,k , the column vector dimensions after this interval are the same as the number of antennas N of the relay node. The signal form received by the p-th subcarrier above is further simplified as follows: Here, ⊙ is the Hadamard product between vectors or matrices of the same dimension. It can be seen that after forwarding by the relay communication system, the communication link channel influences of the relay node and the transceivers in the two links are multiplicative. This means that in order to ensure the maximum signal-to-noise ratio reception, the system processing weights can be designed to comprehensively consider the influence of the two links of the relay link without the need for segmented processing.

7. The method for sensing wireless space-frequency channels in a distributed multi-node relay communication system based on error feedback according to claim 6, characterized in that: Get two estimates of the "coherent combined channel sum" and the "null suppression channel sum": From the above formula, we can see that h k,pre,∑ The complex amplitude influence of all relay space-frequency channels is included in , so the normalized feedback error is constructed on this basis: Because the above formula is taken from the channel estimation results of two consecutive training sequences, when the receiving end is well synchronized, the influence of the initial phase difference of the carrier phase between the transmitter and the receiver is eliminated by dividing each other, and only the role of the relay transceiver channel is reflected in it; After the wireless receiver demodulates and calculates the feedback error, the error e is fed back to all relay nodes through the signaling control link or other wired or wireless communication channels. k .

8. The method for sensing wireless space-frequency channels in a distributed multi-node relay communication system based on error feedback according to claim 1, characterized in that: Step 5 specifically includes: Each wireless relay node in the system receives the normalized error feedback e from the wireless receiver. k After that, the previous space-frequency coherent combining weights and null suppression weights are iteratively updated, and the process is shown in the following formula: Wherein, μ is the step coefficient of the system's iterative perception of the wireless space-frequency channel, which is usually less than 1. Here, μ=0.5; v k+1 If it is the mod(k+1,NPM)th column of the codebook matrix T, and ||t mod(k+1,NPM) ||=1; repeat steps 2 to 5 according to the above formula to obtain the gradually convergent and stable coherent combining and null-steering suppression space-frequency beam weights, and each relay node only needs to take w k and v k The element corresponding to the node number can be used to weighted forward the received signal; For w k and v k It is presented in the form of space-frequency weights of all relay nodes in the entire system. By configuring the initial state in step 1 and the constraints of each node number, each relay node can obtain the minimum mean square error weighting coefficient of the entire system for the wireless space-frequency channel without interaction.

9. The method for sensing wireless space-frequency channels in a distributed multi-node relay communication system based on error feedback according to claim 1, characterized in that: Step 6 specifically includes: As the number of iterations of the transmission and reception communication and feedback error weights increases, the coherent combination weights of the distributed multi-node relay system will gradually converge to the space-frequency channel h according to the minimum mean square error direction. (t) ⊙h (r) In the same direction, the transmit and receive channels of the system relay link can be expressed by the following formulas: When the iterative error has converged to a small value, the currently converged coherent combining weight w is sent to the wireless transmitter or receiver through any distributed relay node. k , so that the transceiver node can obtain the current wireless space-frequency transmission channel h (t) ⊙h (r) Perceptual estimation.

10. The method for sensing wireless space-frequency channels in a distributed multi-node relay communication system based on error feedback according to claim 9, characterized in that: When the transmitter knows the position and attitude of its own transmitting antenna and the antenna arrays of each distributed node, h (r) The transmitter is known, and the angle and delay of the target receiving device relative to the mth distributed node are obtained by the following formula: Among them, a m (θ) and g(τ) are the spatial steering vector and frequency domain channel response caused by the time delay τ of the mth distributed relay node, respectively. Based on the position of each relay node, the coordinate position (x r ,y r ,z r ).

Images