A full-duplex decode-and-forward relay transmission system based on cooperative space-time coding

Abstract

The application discloses a full-duplex decode-and-forward relay transmission system based on cooperative space-time coding and belongs to the technical field of wireless transmission. The system comprises a sending end, a receiving end and a relay end. The sending end comprises a source node, which transmits data after short convolution error correction coding, modulation and composition of asynchronous cooperative frames. The relay end comprises a plurality of relay nodes, each of which is in a simultaneous same-frequency full-duplex state, adopts a decode-and-forward mode, decodes the received signal transmitted by the source node, re-codes and modulates the signal, completes asynchronous cooperative space-time coding, realizes cooperative relay transmission, and the receiving end comprises a destination node which receives the combined signal of the direct link from the source node to the destination node and the relay link from each relay node to the destination node, obtains the equivalent channel information of the signal by using the orthogonality of CHU sequences, and performs low-complexity MMSE linear space-time equalization and demodulation decoding on the received signal to recover the transmitted data of the source node.

Description

TECHNICAL FIELD

[0001] The application relates to a full-duplex decode-and-forward relay transmission system based on cooperative space-time coding and belongs to the technical field of radio transmission. BACKGROUND

[0002] Full-duplex relay technology is generally used to improve the transmission performance of a wireless communication system and expand the communication coverage. Compared with half-duplex relay transmission, full-duplex relay transmission can theoretically increase the throughput of the system by 1 times, but full-duplex relay transmission also causes relay loop self-interference. In recent years, many scholars have conducted a large amount of research on the self-interference and cooperative transmission of full-duplex relay systems. Through self-interference cancellation in the spatial domain, the analog domain and the digital domain, the relay self-interference can be suppressed to be small enough in some scenarios, so that full-duplex relay transmission can be realized. Full-duplex relay modes mainly include amplify-and-forward and decode-and-forward modes. The decode-and-forward protocol decodes the received signal at the relay node, then re-encodes and modulates the new signal obtained after decoding and forwards it, and can suppress the loop self-interference below the noise.

[0003] Distributed space-time coding is an idea of applying space-time coding to a cooperative communication environment and introducing time and space correlation into signals transmitted by distributed antennas. Distributed space-time coding technology mainly includes synchronous distributed space-time codes and asynchronous distributed space-time codes. Asynchronous distributed space-time coding can tolerate the relative time delay between relay nodes and has been proved to be an effective way to obtain cooperative diversity in half-duplex relay networks, but these space-time coding schemes face new problems when applied to high-speed air self-organizing networks. Mainly including the Doppler shift problem caused by high-speed self-organizing networks, the asynchronous space-time cooperative coding problem and the problem of obtaining full diversity gain with low complexity.

[0004] In summary, full-duplex relay transmission technology has greater potential than half-duplex relay transmission technology in improving spectral efficiency, but for the problem of full-duplex relay transmission in the high-speed air self-organizing network scene, the existing technical solutions cannot meet the networking requirements. SUMMARY

[0005] The technical problem solved by the application is to overcome the shortcomings of the prior art and provide a full-duplex decode-and-forward relay transmission system based on cooperative space-time coding, which is used for realizing full-duplex cooperative relay transmission with high spectral efficiency, good anti-interference, anti-noise performance and anti-Doppler frequency shift and applied to full-duplex self-organizing networks in a high-speed flight state.

[0006] The application solves the above technical problems by the following technical scheme:

[0007] A full-duplex decode-and-forward relay transmission system based on cooperative space-time coding comprises:

[0008] The transmitting end includes a source node, which performs short convolutional error correction coding, modulation, and asynchronous cooperative frame sequence on the raw bit stream of the data to be transmitted before transmission;

[0009] The relay end includes several relay nodes, each of which is in a simultaneous, same-frequency, full-duplex state. Each relay node adopts a decode-forward mode, decodes the received signal transmitted by the source node, and then re-encodes, modulates, asynchronously cooperates with space-time coding, forms an asynchronous cooperative frame sequence and inserts a channel estimation sequence, and then sends the signal to the destination node.

[0010] The receiving end includes a destination node, which is used to receive the combined signals of the direct link from the source node to the destination node and the relay link from the relay node to the destination node, obtain the equivalent channel information of the above signals, and then perform linear space-time equalization, demodulation and decoding on the signals to finally obtain the data transmitted by the source node.

[0011] Preferably, the asynchronous cooperative frame consists of a training sequence and multiple data symbol subframes encoded by short convolutional error correction, N fb N represents the number of symbols contained in a data symbol subframe; the leading edge of the data symbol subframe is a zero guard interval, and the length of the zero guard interval is N. z Not less than the maximum link delay difference of the direct link and each relay link; a channel estimation sequence is inserted after every N asynchronous cooperative frames, where the value of N is determined by the effective data transmission rate requirement, the length of the training sequence, and the length of the channel estimation sequence.

[0012] Preferably, the Kth relay node forms an asynchronous cooperative frame sequence, and the training sequence in the asynchronous cooperative frame consists of two identical CHU sequences x. chup_K And composed of zero protection interval, x chup_K Length N fb -N z The channel estimation sequence for the Kth relay node consists of two identical CHU sequences x. chues_K Composition, x chues_K The length is L*N fb L > 1; the training sequences of all different relay nodes are mutually orthogonal, and the channel estimation sequences of all different relay nodes are mutually orthogonal.

[0013] Preferably, the decoding of the relay node includes synchronization, Doppler frequency offset estimation of the training sequence and the channel estimation sequence, channel estimation, frequency offset compensation, channel equalization, frequency offset correction and phase compensation based on decision feedback iteration, and decoding.

[0014] Preferably, the joint Doppler frequency offset estimation of the training sequence and the channel estimation sequence includes: performing correlation accumulation on the training sequence and the channel estimation sequence in the synchronized source node transmitted signal with different window lengths, taking the phase of the two accumulation results respectively to obtain frequency offset estimation results f1 and f2, and weighting the two frequency offset estimation results to obtain the final frequency offset estimation result.

[0015] Preferably, the frequency offset correction and phase compensation based on decision feedback iteration includes:

[0016] 1) Remapping: Equalizing the symbol y′ of the i-th data symbol subframe k Demap and remap as known signals to constellation points s′. k ;

[0017] 2) Channel re-estimation: via symbol y′ k The remapped symbol s′ k Channel estimation is performed, and the following is obtained:

[0018]

[0019] In the formula, k represents the k-th symbol in the i-th data symbol subframe. The average channel estimation bias for the i-th data symbol subframe;

[0020] The average deviation of the obtained channel estimation Take the phase to obtain the overall remaining rotation phase of the i-th data symbol subframe.

[0021] 3) Phase Iteration Compensation: The overall residual rotating phase is compensated into the channel estimation result to obtain a new channel estimation result. For channel equalization used in the (i+1)th data symbol subframe;

[0022] 4) Residual frequency offset estimation: Equalize the symbol y′ k With remapping symbol s′ k Perform conjugate multiplication, take the phase, and then sum them to obtain the residual frequency offset sum fdr_sum for this subframe;

[0023] The channel equalization of the i-th data symbol subframe uses the channel estimation result corrected in the (i-1)-th data symbol subframe. Starting with the first symbol of the (i-1)th data symbol subframe, and assuming the unit residual frequency offset is fdr, then the residual frequency offset of each symbol in the (i-1)th data symbol subframe is 0, fdr, 2fdr, ..., (N). fb -1)fdr, the residual frequency offset of each symbol in the i-th data symbol subframe is N. fb ·fdr、(Nfb +1) · fdr,..., (2N fb -1) · fdr, then the total number of the phase-accumulated residual frequency offset of the ith data symbol sub-frame is:

[0024]

[0025] The average residual frequency offset fdr of the ith data symbol sub-frame is calculated i ;

[0026] 5) Frequency offset iteration: the frequency offset estimation value fd of the modified Doppler frequency offset estimation by the average residual frequency offset i , then fd i+1 = fd i + fdr i , which is used for frequency offset compensation of the (i+1)th data symbol sub-frame.

[0027] Preferably, the asynchronous cooperative space-time coding adopts a convolutional encoder with several delay-weighted branches, and the weighting coefficients are determined by the encoding generation matrix of the asynchronous cooperative space-time coding.

[0028] Preferably, the asynchronous cooperative space-time coded signal is t K = A K s, is a certain data symbol sub-frame forwarded by the relay node, is the encoding generation matrix of the Kth relay node, and q = b + N fb -1, b is the convolution length, and A K is a Toeplitz matrix:

[0029]

[0030] If the asynchronous cooperative space-time coding of the relay node is implemented by a convolutional encoder with several delay-weighted branches, then the encoding vector of the Kth relay node is m k = [v1, v2,..., v b ], v1, v2,..., v b are the weighting coefficients of the convolutional encoder for implementing the asynchronous cooperative space-time coding of the relay node, the total number of the relay nodes is r, m0 represents the direct link and is a row vector of 1*b [1, 0,..., 0], and the asynchronous cooperative space-time coding matrix M is composed of the direct link and the encoding vectors of all the relay nodes:

[0031]

[0032] The encoding matrix M is designed based on the translation full-rank matrix criterion to obtain full diversity gain.

[0033] Preferably, the channel estimation sequence received by the destination node is:

[0034]

[0035] wherein y0 is the channel estimation sequence sent by the source node, y1, …, y K … are the channel estimation sequences sent by all relay nodes, n is the noise; K represents the Kth relay node, K is less than or equal to the number of all relay nodes;

[0036] By using the orthogonality of the channel estimation CHU sequences among different relay nodes, the received channel estimation signals are multiplied by the conjugate transpose matrices of the channel estimation sequences respectively, and divided by the autocorrelations of the channel estimation sequences, so as to obtain the direct link channel estimation results from the source node to the destination node Channel estimation results of the links from the relay nodes to the destination node

[0037] The obtained equivalent channel information is:

[0038]

[0039] wherein A0, A1, …, A K … are the encoding generation matrices of the source node and the relay nodes respectively, is the equivalent channel information obtained by the destination node.

[0040] Preferably, after the equivalent channel information is obtained, the MMSE channel estimation algorithm is used to perform linear space-time equalization on the signals.

[0041] Compared with the prior art, the present application has the following advantages:

[0042] (1) Compared with the existing half-duplex relay transmission mode, the present application innovatively proposes an asynchronous cooperative space-time coding transmission mode, which can multiply the frequency offset efficiency while obtaining full diversity gain. The zero guard interval design of the asynchronous cooperative frame structure effectively avoids the inter-symbol interference between the symbol frames caused by the signal arrival delay; the convolution error correction code design can effectively reduce the demodulation delay and reduce the relay forwarding delay; the channel estimation sequence and training sequence structure design can effectively improve the Doppler shift problem caused by the high-speed ad hoc network.

[0043] (2) The system is a relay cooperative communication, and the relay nodes work in the simultaneous frequency full-duplex state, adopt the decode-and-forward mode, perform self-interference cancellation to the noise level on the received signals, and reduce the influence of the self-interference on the effective signals received by the relay as much as possible, so as to expand the transmission distance.

[0044] (3) The relay node and the destination node adopt the Doppler frequency offset estimation method of training sequence and channel estimation sequence, which reduces the frequency offset estimation error caused by insufficient training sequence length and improves the noise resistance and estimation accuracy of frequency offset estimation.

[0045] (4) The relay node and the destination node adopt a frequency offset correction and compensation method based on decision feedback iteration, which improves the frequency offset estimation accuracy, reduces the negative impact of the accumulation of residual frequency offset estimation error on the demodulation result, and solves the Doppler frequency offset compensation problem under high-speed flight conditions.

[0046] (5) Asynchronous cooperative space-time coding is performed between relay nodes or between relay nodes and direct links. Low-complexity MMSE linear space-time equalization is used at the destination node, which can obtain full cooperative diversity gain at the receiving end. Attached Figure Description

[0047] Figure 1 A schematic diagram of a full-duplex decoding-forwarding relay transmission system based on asynchronous cooperative space-time coding provided in an embodiment of the present invention;

[0048] Figure 2 This is a schematic diagram of the asynchronous cooperative frame structure provided in an embodiment of the present invention;

[0049] Figure 3 This is a flowchart of relay node signal processing provided in an embodiment of the present invention;

[0050] Figure 4 A schematic diagram of the asynchronous cooperative space-time coding principle provided in an embodiment of the present invention;

[0051] Figure 5 The bit error rate simulation comparison chart provided for embodiments of the present invention is a comparison chart of the traditional transmission mode. Detailed Implementation

[0052] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0053] The effects of the present invention will be described in detail below with reference to simulation.

[0054] like Figure 1 As shown, a node has a source node S, relay nodes R1, R2...R... K Taking the simultaneous full-duplex relay decoding and forwarding relay cooperative communication model of destination node D as an example, the relay node adopts the simultaneous full-duplex and decoding forwarding mode, and the destination node receives the direct link signal from the source to the destination node and the relay link signal from the relay node to the destination node.

[0055] likeFigure 2 As shown, the asynchronous cooperative frame structure consists of a length of 2N. fb The training sequence and its length are N f ·N fb The data symbol frames composed of short convolutional error correction codes (N) fb The number of symbols contained in a data symbol subframe, including N. z (Number of zero-guard intervals), the length of which is designed to be no less than the maximum link delay difference between the source and destination nodes. An insertion length of 2L*N is inserted after every N asynchronous cooperative frames. fb The channel estimation sequence is L>1. The training sequence consists of two identical sequences of length N. f -N z CHU sequence x chup_K The training sequence consists of a zero guard interval and a total length of 2N. fb The channel estimation sequence consists of two identical sequences of length L*N. fb CHU sequences x with L > 1 chues_K The total length of the channel estimation sequence is 2L*N. fb ,L>1.

[0056] like Figure 3 As shown, the decoding process of a relay node includes synchronization, joint Doppler frequency offset estimation of the training sequence and channel estimation sequence, channel estimation, frequency offset compensation, equalization, decoding, and remapping, residual frequency offset estimation, channel re-estimation, frequency offset iteration, and phase iteration compensation in frequency offset correction and phase compensation based on decision feedback iteration. The forwarding process of a relay node includes short convolutional error correction coding, mapping, asynchronous cooperative space-time coding, and framing. The specific steps of the joint Doppler frequency shift estimation method of the training sequence and channel estimation sequence are as follows:

[0057] 1) Apply a window of length N to both the training sequence and the channel estimation sequence. fb and L*N fb The delayed correlation accumulation for L>1 yields cor_sum1 and cor_sum2:

[0058]

[0059] Where y i (k) represents the received training sequence or channel estimation sequence, W i Indicates the window length.

[0060] 2) Take the phase of cor_sum1 and cor_sum2 respectively, and obtain two frequency offset estimation results f1 and f2:

[0061]

[0062] 3) Weighting the two frequency offset estimation results to obtain the final frequency offset estimation result:

[0063]

[0064] The specific steps of the frequency offset correction and phase compensation method based on decision feedback iteration are as follows:

[0065] 1) Re-mapping: Re-mapping the symbols y' after equalization of the i-th data symbol sub-frame k as known signals into constellation points s' k .

[0066] 2) Channel re-estimation: Channel estimation is performed on the symbols y' after equalization k and the symbols s' after re-mapping k to obtain:

[0067]

[0068] where k represents the k-th symbol in the i-th data symbol sub-frame, is the average value of the channel estimation bias of the i-th data symbol sub-frame, and the overall residual rotation phase of the i-th data symbol sub-frame can be obtained by taking the phase of

[0069]

[0070] 3) Phase iterative compensation: The estimated overall residual rotation phase is compensated into the channel estimation result to obtain a new channel estimation result for channel equalization of the i+1-th data symbol sub-frame.

[0071]

[0072] 4) Residual frequency offset estimation: Parallel to steps 2) and 3), the symbols y' after equalization k and the symbols s' after re-mapping k are conjugate multiplied, and the residual frequency offset accumulation of the sub-frame is obtained by taking the phase and accumulating:

[0073]

[0074] The channel equalization of the i-th data symbol sub-frame uses the channel estimation result corrected in the i-1-th data symbol sub-frame Therefore, taking the 1st symbol of the i-1-th data symbol sub-frame as the starting point, and assuming that the unit residual frequency offset is fdr, then the residual frequency offset of each symbol of the i-1-th data symbol sub-frame is 0, fdr, 2fdr, …, (N fb-1)fdr, the residual frequency offset of each symbol of the i-th data symbol subframe is N fb · fdr, (N fb +1)· fdr, …, (2N fb -1)· fdr, the total number of residual frequency offsets after phase taking of the i-th data symbol subframe is:

[0075]

[0076] The average residual frequency offset estimated for the i-th data symbol subframe is:

[0077]

[0078] 5) Frequency offset iteration: using the residual frequency offset to correct the frequency offset estimation value: fd i+1 = fd i + fdr i , and is used for frequency offset compensation of the i+1 data symbol subframe.

[0079] As shown in Figure 4 , the transmitted signal after asynchronous cooperative space-time coding is t K = A K s. A K is a Toeplitz matrix K = A K s, is a certain data symbol subframe forwarded by the relay node, is the encoding generation matrix of the K-th relay node, and q = b + N fb -1, b is the convolution length, and A K is a Toeplitz matrix (a Toeplitz matrix is obtained by cyclically shifting the same vector, and the convolution length is b), and the specific form is as follows:

[0080]

[0081] If the asynchronous cooperative space-time coding of the relay node is implemented by using a convolutional encoder with a plurality of delay weighted branches, the encoding vector of the K-th relay node is m k = [v1, v2, …, v b ], v1, v2, …, v b are the weighted coefficients of the convolutional encoder for implementing the asynchronous cooperative space-time coding of the relay node, the total number of relay nodes is r, m0 represents the direct link, and is a row vector of 1*b [1, 0, …, 0]. The encoding matrix is composed of the direct link and the encoding vectors of all relay nodes:

[0082]

[0083] The asynchronous cooperative space-time coding matrix is designed based on the shift full rank (SFR) criterion to obtain full diversity gain.

[0084] The destination node receives the combined signal of the direct link of the source-destination node and the relay link of each relay-destination node, and the received channel estimation sequence is:

[0085]

[0086] In the formula, y0, y1, …, y K … are the channel estimation sequences sent by the source node and each relay node respectively;

[0087] By using the orthogonality of the channel estimation CHU sequences between different relay nodes, the received channel estimation signal is multiplied by the conjugate transpose matrix of each channel estimation sequence respectively, and divided by the autocorrelation of each channel estimation sequence, to obtain the channel information of different links respectively, and the direct link is as follows:

[0088]

[0089] For the Kth relay node, there is:

[0090]

[0091] The obtained equivalent channel information is:

[0092]

[0093] In the formula, A0, A1, … A K … are the encoding generation matrices of the source node and each relay node respectively, are the channel estimation results of the source-destination node link and each relay-destination node link respectively, is the equivalent channel information obtained by the destination node.

[0094] Then the equivalent channel matrix is Then the MMSE linear space-time equalization is carried out by using the MMSE channel estimation algorithm:

[0095]

[0096] In the formula, σ 2 is the noise variance, I K is the unit matrix. fb *N fb .

[0097] The technical effects of the present application are further described in combination with simulation experiments:

[0098] 1) Simulation conditions

[0099] The simulation uses Matlab R2021a simulation software. One data symbol subframe contains N fb = 33 symbols, one asynchronous cooperation frame contains N f = 20 data symbol subframes, the zero interval length N z = 3, and channel estimation is performed once every N = 5 asynchronous cooperation frames. The baseband symbol rate is 250 kHz, and the maximum frequency offset is 250 Hz. The channel is a Rayleigh fading channel, and for a full-duplex decode-and-forward relay cooperation model including three communication nodes of a source node, a relay node and a destination node, a difference of 6 dB between a source-relay node link, a relay-destination node link and a source-destination node link is considered.

[0100] An asynchronous cooperation space-time coding matrix is designed based on the shift full rank matrix (SFR) criterion to obtain full diversity gain. For the full-duplex decode-and-forward relay cooperation model of the above three communication nodes, the coding matrix satisfying the shift full rank matrix (SFR) criterion is designed as

[0101]

[0102] Then the asynchronous cooperation space-time coding of the relay node corresponds to a convolutional encoder with three delay-weighted branches, and the weighting coefficients are

[0103] 2) Simulation content and simulation results

[0104] Three transmission modes of decode-and-forward (DF) without asynchronous cooperation space-time coding, amplify-and-forward (AF) with asynchronous cooperation space-time coding, and direct transmission from the source node to the destination node are used as comparison methods, and the proposed method is simulated and compared with the three comparison methods, and the results are shown in Figure 5 The simulation results show that the simultaneous frequency full-duplex decode-and-forward relay system based on asynchronous cooperation space-time coding has better system performance and can obtain full diversity gain.

[0105] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. Any modification, equivalent replacement and improvement made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A full-duplex decode-and-forward relay transmission system based on cooperative space-time coding, characterized in that, The application relates to a method for transmitting data in a cooperative communication system. The transmitting end comprises a source node for transmitting a data original bit stream to perform short convolution error correction coding, modulation, asynchronous cooperation frame sequence composition and transmission; The relay end comprises a plurality of relay nodes, each of which is in a simultaneous same-frequency full-duplex state; each relay node adopts a decode-and-forward mode to decode the source node transmission signal received, and then re-performs coding, modulation, asynchronous cooperation space-time coding, asynchronous cooperation frame sequence composition and channel estimation sequence insertion, and sends the signal to a destination node; The receiving end comprises a destination node for receiving a combined signal of a direct link from the source node to the destination node and a relay link from the relay node to the destination node, obtaining equivalent channel information of the signal, performing linear space-time equalization and demodulation and decoding on the signal, and finally obtaining the transmission data of the source node; The asynchronous cooperative frame is composed of a training sequence and a plurality of data symbol subframes after short convolution error correction coding, N fb The number of symbols contained in one data symbol subframe; A zero protection interval is arranged at the front end of a data symbol subframe, and the duration of the zero protection interval is not less than the maximum link delay difference of the direct link and the relay links; a channel estimation sequence is inserted after every N asynchronous cooperation frames, and the value of N is determined by the effective data transmission rate requirement, the training sequence length and the channel estimation sequence length; The Kth relay node constitutes an asynchronous cooperation frame sequence, and the training sequence in the asynchronous cooperation frame is composed of two identical CHU sequences x chup_K and a zero guard interval, x chup_K with a length of N fb -N z , N z is the length of the zero guard interval; the channel estimation sequence of the Kth relay node is composed of two identical CHU sequences x chues_K , x chues_K with a length of L*N fb , L>1; the training sequences of all different relay nodes are mutually orthogonal, and the channel estimation sequences of all different relay nodes are mutually orthogonal.

2. The cooperative space-time coding based full-duplex decode-and-forward relay transmission system according to claim 1, wherein, The decoding of the relay node comprises synchronization, joint Doppler frequency offset estimation of a training sequence and a channel estimation sequence, channel estimation, frequency offset compensation, channel equalization, frequency offset correction and phase compensation based on decision feedback iteration and decoding.

3. The cooperative space-time coding based full-duplex decode-and-forward relay transmission system according to claim 2, wherein, The joint Doppler frequency offset estimation of the training sequence and the channel estimation sequence comprises: correlation accumulation of the training sequence and the channel estimation sequence in the synchronized source node transmission signal is respectively performed with different window lengths, the phases of the two accumulation results are respectively taken to obtain frequency offset estimation results f1 and f2, and the final frequency offset estimation result is obtained by weighting the two frequency offset estimation results.

4. The cooperative space-time coding based full-duplex decode-and-forward relay transmission system according to claim 2, wherein, The frequency offset correction and phase compensation based on decision feedback iteration comprises: 1) remap: remap the equalized symbols y' of the ith data symbol subframe k demap, and remap as known signal into constellation point s k ′; 2) Channel re-estimation: by the symbol y' k , the remapped symbol s' k Channel estimation is performed to obtain: where k represents the kth symbol in the ith data symbol subframe, is the channel estimation bias average for the ith data symbol subframe. averaging the channel estimation bias obtained taking the phase, obtaining the overall residual rotational phase of the i-th data symbol subframe 3) Phase iterative compensation: compensate the overall residual rotation phase into the channel estimation result to get a new channel estimation result for channel equalization for the i+1th data symbol subframe; 4) residual frequency offset estimation: multiply the equalized symbol y' with the remapped symbol s' k k and take the phase to get the residual frequency offset sum fdr_sum for this subframe.​ The channel equalization of the ith data symbol sub-frame uses the channel estimation result corrected by the (i-1)th data symbol sub-frame Taking the first symbol of the (i-1)th data symbol sub-frame as a starting point, assuming that the unit residual frequency offset is fdr, then the residual frequency offset of each symbol of the (i-1)th data symbol sub-frame is 0, fdr, 2fdr, …, (N fb -1)fdr, the residual frequency offset of each symbol of the ith data symbol sub-frame is N fb ·fdr, (N fb +1)·fdr, …, (2N fb -1)·fdr, then the total number of residual frequency offsets after phase accumulation of the ith data symbol sub-frame is: The average residual frequency offset fdrfor the ith data symbol subframe estimate is computed as i ; 5) Frequency offset iteration: correct the frequency offset estimate fd obtained from the average residual frequency offset correction of the Doppler frequency offset estimate i then fd i+1 = fd i + fdr i for frequency offset compensation of the i+1 data symbol subframe.

5. The cooperative space-time coding based full-duplex decode-and-forward relay transmission system according to claim 1, wherein, The asynchronous cooperation space-time coding adopts a convolutional encoder with a plurality of delay weighting branches, and the weighting coefficients are determined by a coding generation matrix of the asynchronous cooperation space-time coding.

6. The cooperative space-time coding based full-duplex decode-and-forward relay transmission system according to claim 5, wherein, Asynchronous cooperative space-time coded signal for t K = A K s, is a certain data symbol subframe forwarded by the relay node, is the encoding generator matrix for the Kth relay node, and where q = b + N fb -1, b is the convolution length, A K is a Toeplitz matrix: The asynchronous cooperative space-time coding of the relay nodes is implemented by using a convolutional encoder with several delay-weighted branches, and the coding vector of the Kth relay node is m k = [v1, v2, …, v b ], v1, v2, …, v b , which are the weight coefficients of the convolutional encoder for implementing the asynchronous cooperative space-time coding of the relay node. Assuming that the total number of the relay nodes is r, m0 represents the direct link and is a row vector of 1*b [1, 0, …, 0], and the asynchronous cooperative space-time coding matrix M is composed of the direct link and the coding vectors of all the relay nodes: The coding matrix M is designed based on a shift full-rank matrix criterion to obtain full diversity gain.

7. The cooperative space-time coding based full-duplex decode-and-forward relay transmission system according to claim 6, wherein, The channel estimation sequence received by the destination node is: where y0is the channel estimation sequence transmitted by the source node, y1,..., y K ... are the channel estimation sequences transmitted by all relay nodes, n is the noise; K represents the Kth relay node, K is less than or equal to the number of all relay nodes; The orthogonality of the channel estimation CHU sequences between different relay nodes is utilized, the received channel estimation signals are multiplied by the conjugate transpose matrices of the channel estimation sequences respectively, and are divided by the autocorrelations of the channel estimation sequences, to obtain the direct link channel estimation results from the source node to the destination node respectively The channel estimation results of the links from the relay nodes to the destination node The obtained equivalent channel information is: In the formula, A0, A1, … A K … are the encoding generation matrices of the source node and the respective relay nodes, respectively, is the equivalent channel information obtained for the destination node.

8. The cooperative space-time coding based full-duplex decode-and-forward relay transmission system of claim 7, wherein, After the equivalent channel information is obtained, the MMSE channel estimation algorithm is adopted to perform linear space-time equalization on the signal.

Images