Weighted update mechanism-based MRC signal detection method and apparatus

Abstract

The invention discloses an MRC signal detection method and device based on a weighted update mechanism, mainly solving the problem that the existing MRC detection method is poor in performance in a high-order modulation mode of a coding OTFS system, and the implementation scheme comprises the following steps: after LDPC coding, interleaving, modulation and mapping are carried out on a sending bit sequence at a sending end, a time delay-Doppler DD domain sending signal is generated, and the time delay-Doppler DD domain sending signal is sent to the sending end; converting the signal into a time domain sending signal suitable for channel transmission; the receiving end reconstructs the time domain receiving signal into a time delay-time (DT) domain receiving signal; then MRC iteration detection based on a weighted updating mechanism is carried out on a DT domain receiving signal to obtain a DT domain sending signal vector estimation value, and LLR conversion, de-interleaving and LDPC decoding judgment are sequentially carried out after the DT domain sending signal vector estimation value is converted into a DD domain to obtain an estimation value of a sending message bit sequence. According to the method, the detection result and the decoding result are weighted and combined by setting a proper weight factor, so that the overall performance of the system is improved, and the method can be used for a coding OTFS system in a high-speed moving scene.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communication, and particularly relates to an MRC signal detection method and device, which can be used in a coded OTFS system in a high-speed mobile scenario. Background Art

[0002] The sixth-generation mobile communication 6G system is expected to support a large number of high-frequency communication mobile terminals, such as drones, high-speed railways, low-earth orbit satellites, and highway vehicles. High spectral efficiency, high energy efficiency, and low-latency reliable communication become more important. In the above scenarios, the relative speed between the transceiver ends may be hundreds of kilometers per hour, and the channel state information changes rapidly. The orthogonal frequency division multiplexing OFDM technology widely used in 4G and 5G faces severe inter-carrier interference in a high-speed mobile environment and is difficult to provide stable and reliable communication services. Orthogonal time-frequency-space OTFS modulation, with its characteristic of adapting to fast time-varying wireless channels, has become one of the candidate technologies for 6G. The OTFS technology introduces the delay-Doppler DD domain through a two-dimensional Fourier transform on the basis of the OFDM technology. The transmitting end distributes each information symbol in the DD domain, converts the fast-fading time-varying channel in the time-frequency TF domain into a quasi-static time-invariant channel in the DD domain, so as to combat the Doppler frequency shift and achieve reliable communication.

[0003] Currently proposed OTFS signal detection methods include the minimum mean square error LMMSE algorithm, the message passing MP and improved algorithms, the maximum ratio combining MRC, and the signal detection algorithm based on deep learning, etc. However, these detection algorithms are mainly optimized for uncoded OTFS systems and do not fully consider the potential performance improvement of coded OTFS systems, resulting in limitations in the anti-interference ability and bit error rate of uncoded OTFS systems. In order to further improve the system performance, some studies have combined channel coding with OTFS detection algorithms to enhance the anti-interference ability of the system and reduce the bit error rate, thereby improving the overall communication reliability.

[0004] In 2020, Thaj T, Viterbo E et al. combined the MRC detection algorithm with the low-density parity-check code LDPC encoding and decoding technology in their paper "Low Complexity Iterative Rake Decision Feedback Equalizer for Zero-Padded OTFS Systems" (IEEE Transactions on Vehicular Technology, 2020) to further improve the system performance. However, in each iteration of this method, regardless of the result of the MRC detector, the initial symbol estimation in the next iteration completely depends on the output of the LDPC decoder. If the output of the LDPC decoder is unstable or has a large error, these unreliable initial estimates may accumulate continuously during subsequent iterations, ultimately leading to a decline in the system bit error performance.

[0005] In 2021, Li S Y, Yuan J H and Yuan W J extended the uncoded OTFS system to the coded OTFS system and analyzed the performance of the coded OTFS system in their paper "Performance Analysis of Coded OTFS Systems over High-Mobility Channels" (IEEE Transactions on Wireless Communications, 2021). However, they mainly focused on the relationship between the coding gain and the diversity gain in OTFS modulation and did not deeply explore the performance of the detection algorithm under the coded system.

[0006] The patent document with the application number CN202411010293.1 discloses an iterative decoding feedback detection method for an underwater acoustic OTFS communication system. It constructs an outer-loop iterative structure between the DD-domain equalizer and the LDPC decoder to improve the system performance through multiple iterative feedbacks. However, since this method does not clearly specify the specific type of the equalizer and only conducts simulation analysis for low-order modulation without considering the performance under high-order modulation, its applicability and flexibility in the actual system will be limited. At the same time, because the prior information of the DD-domain equalizer completely depends on the output of the LDPC decoder and lacks the utilization of other information, the performance of the equalizer is directly affected by the output quality of the LDPC decoder, resulting in a decline in the overall communication performance. Summary of the Invention

[0007] The purpose of the present invention is to address the above deficiencies of the existing technologies and propose an MRC signal detection method and device based on a weighted update mechanism to achieve more efficient signal detection for the coded OTFS system and improve the bit error performance of the system in a high-speed mobile environment.

[0008] The technical idea of the present invention is that during the MRC detection iteration process, by setting an appropriate weighting factor, the MRC detection result and the decoding result are combined with weights, and the weighted result is used as the initial symbol estimation for the next iteration of the MRC detector. In this way, the reliable information after decoding can be used to optimize the detection result, and the soft value estimation of the current iteration can be retained, avoiding complete dependence on the output of the LDPC decoder. Through this balancing mechanism, the MRC soft value output is coordinated with the LDPC decoding result, thereby improving the overall performance of the system and achieving better bit error rate performance.

[0009] According to the above idea, the technical solution of the present invention includes the following:

[0010] 1. An MRC signal detection method based on a weighted update mechanism, characterized by comprising:

[0011] (1) At the transmitting end, the binary transmitted message bit sequence u is encoded by a low-density parity-check code LDPC, interleaved, modulated and mapped to generate a delay-Doppler DD-domain transmitted signal X. The DD-domain transmitted signal X is converted to obtain a time-domain transmitted signal z and transmitted through a wireless channel;

[0012] (2) At the receiving end, the time-domain received signal r is matrix-reconstructed to obtain a delay-time DT-domain received signal Y;

[0013] (3) Using a Turbo iterative structure, perform MRC iterative detection based on a weighted update mechanism on the DT-domain received signal That is, MRC detector, LLR converter, deinterleaving, LDPC decoder, interleaving, QAM modulation and weighted update, to obtain the DT-domain transmitted symbol vector estimation value

[0014]

[0015] where m is the row index of the DD-domain grid, δ is the weighting factor, 0 < δ ≤ 1; is the DT-domain MRC soft value estimation vector output by the MRC detector, is the reconstructed modulation vector output after QAM modulation;

[0016] (4) Convert the DT-domain transmitted symbol vector estimation value to the DD-domain transmitted symbol vector estimation value and perform LLR conversion, deinterleaving, and LDPC decoding decision on in sequence to obtain the estimated value

[0017] 2. The MRC signal detection device based on the weighted update mechanism is characterized by comprising:

[0018] The MRC detector module is used to extract and combine the received multipath components of the transmitted symbols based on maximum ratio combining in the DT domain, improve the signal-to-noise ratio of the combined signal, enhance the reliability of the signal, and provide a more accurate input for the subsequent bit-level log-likelihood ratio (LLR) calculation;

[0019] The LLR converter module is used to convert the soft value information output by the MRC detector into the bit-level log-likelihood ratio (LLR) to provide reliability information for the subsequent channel decoding;

[0020] The deinterleaving module is used to restore the bit-level LLR information rearranged by the interleaver to the original order;

[0021] The LDPC decoder module is used to perform LDPC decoding operation on the data after deinterleaving to obtain hard decision encoded bits;

[0022] The interleaving module is used to perform interleaving processing on the hard decision encoded bits output by the LDPC decoder, that is, by rearranging the order of the data, so that the consecutive errors occurring during the transmission are dispersed, and the resistance of the system to burst errors is enhanced;

[0023] The QAM modulation module is used to perform QAM modulation on the interleaved data again to obtain a reconstructed modulation vector containing decoding information;

[0024] The weighted update module is used to perform weighted combination of the soft value information output by the MRC detector and the reconstructed modulation signal output by the QAM modulation to effectively enhance the interaction information between detection and decoding and achieve better bit error rate performance.

[0025] Compared with the prior art, the present invention has the following advantages:

[0026] By setting appropriate weight factors, the present invention performs weighted combination on the MRC detection result and the LDPC decoding result, and uses the weighted result as the initial symbol estimation value for the next iteration of the MRC detector. Compared with the existing MRC detection, the method of the present invention can not only use the reliable information after decoding to optimize the detection result, but also retain the soft value estimation of the current iteration, realizing the coordination between the MRC soft value output and the LDPC decoding result, thereby effectively enhancing the interaction information between detection and decoding, improving the bit error rate performance, and having more significant performance gain especially under high-order modulation. Description of the Drawings

[0027] Figure 1 is a schematic flowchart of the implementation process of the MRC signal detection method provided in the first embodiment of the present invention;

[0028] Figure 2 Yes Figure 1 Schematic diagram of the iterative detection implementation process of the MRC signal in the DT domain based on the weighted update mechanism;

[0029] Figure 3 It is a diagram of the MRC signal detection device provided in the second embodiment of the present invention;

[0030] Figure 4 It is a comparison diagram of the bit error rate curves simulated by the present invention with different weight factors δ under the 4-QAM modulation mode;

[0031] Figure 5 It is a comparison diagram of the bit error rate curves simulated by the present invention with different weight factors δ under the 16-QAM modulation mode;

[0032] Figure 6 It is a comparison diagram of the bit error rate curves simulated by the present invention with different weight factors δ under the 64-QAM modulation mode. Detailed implementation manners

[0033] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0034] It should be noted that the step numbers in the specification and claims of the present invention are only for clearly describing the implementation solutions of the present invention for easy understanding, and their sequence numbers are not limited.

[0035] Embodiment 1, MRC signal detection method based on weighted update mechanism:

[0036] Refer to Figure 1 , the implementation of this example includes:

[0037] Step 1, construct the time-delay Doppler DD domain OTFS transmission signal X.

[0038] The orthogonal time-frequency-space OTFS signal refers to modulation in the time-delay Doppler domain, which can effectively cope with multipath effects and Doppler frequency shifts in high-speed mobile scenarios, thereby improving the reliability and spectral efficiency of the communication system.

[0039] The implementation of this step is as follows:

[0040] (1.1) Set up an M×N dimensional DD domain grid and set the symbols in the last M′ rows of the grid to zero, where N is the total number of symbol blocks, M is the total number of subcarriers, and M′≥lmax , l max is the maximum delay tap of the channel;

[0041] (1.2) At the transmitter, the binary message bit sequence u ∈ (0, 1) k is input to the LDPC encoder with code rate R, and the encoded sequence c ∈ (0, 1) L is obtained. Then, the encoded sequence c passes through an interleaver to obtain the interleaved sequence c′, where k is the length of the message sequence and L is the total length after encoding;

[0042] (1.3) After performing Q - order quadrature amplitude QAM modulation on the interleaved sequence c′, it is mapped to the grid information symbols of (M - M′)×N in the DD domain to obtain the two - dimensional transmitted signal X. Among them, the m - th row of X is the transposed vector of the DD - domain transmitted signal vector x m ; represents the M×N - dimensional matrix space, m is the row index of the DD - domain grid, and m = 0, 1,..., M - 1.

[0043] Step 2: Convert the DD - domain transmitted signal X to obtain the time - domain transmitted signal z suitable for wireless channel transmission.

[0044] (2.1) Using the OTFS system model based on the Zak transform, perform the N - point inverse Fourier transform N - IFFT on the DD - domain two - dimensional transmitted signal X to obtain the delay - time DT - domain transmitted signal

[0045]

[0046] where F N is the normalized N - point discrete Fourier transform DFT matrix, and (·) H is the conjugate transpose of the matrix;

[0047] (2.2) Perform matrix column - vectorization on the DT - domain transmitted signal to obtain the time - domain transmitted signal z:

[0048]

[0049] where represents the MN×1 - dimensional matrix space, and vec(·) represents matrix column - vectorization. This signal is then transmitted through the wireless channel.

[0050] In the OTFS system, the operations at the transmitter and receiver are usually implemented on the basis of the OFDM system by adding the inverse symplectic Fourier transform (ISFFT) and the symplectic Fourier transform (SFFT). However, the present invention adopts an OTFS system based on the Zak transform, and the conversion of the signal from the DD domain to the DT domain can be completed in only one step, which greatly reduces the complexity of the system.

[0051] Step 3: Obtain the time-domain received signal r.

[0052] (3.1) Transmit the time-domain transmitted signal z over the time-domain equivalent channel H at the transmitter. The time-domain equivalent channel H is expressed as:

[0053]

[0054] Where represents the MN×MN-dimensional matrix space, P is the number of propagation paths, h q is the channel gain of the q-th path, l q and k q are respectively the delay component and the Doppler frequency shift of the q-th path;

[0055] is a permutation matrix, used to simulate the effect of the delay of the q-th path;

[0056] Δ = diag{α 0 , α 1 , α 2 ,..., α MN-1} is a diagonal matrix, used to simulate the effect of the Doppler frequency shift of the q-th path, α = e j2π / MN , j = 0,..., MN - 1;

[0057] (3.2) Obtain the time-domain received signal r at the receiver:

[0058] r = Hz + w

[0059] Where w is the time-domain noise vector,

[0060] Step 4: Obtain the DT-domain received signal

[0061] The receiver reconstructs the matrix of the time-domain received signal r to obtain the DT-domain received signal

[0062]

[0063] Where It is the inverse operation of vec(·), that is, reconstructing the vector into a matrix with M rows and N columns.

[0064] Step 5: For the received signal in the DT domain Perform MRC iterative detection based on the weighted update mechanism in the DT domain to obtain the estimated value of the transmitted symbol vector in the DT domain

[0065] Refer to Figure 2 , the specific implementation of this step is as follows:

[0066] (5.1) Initialize the estimated value of the transmitted symbol vector in the DT domain at the beginning of the iteration Set the maximum number of iterations ite, and calculate the initial residual noise and interference RNPI vector

[0067]

[0068] where i is the number of iterations, and i = 0 at the beginning of the iteration;

[0069] 0 N is a zero vector of length N;

[0070] is the estimated signal in the DT domain The column vector of the symbol in the m-th row, m = 0,..., M - 1;

[0071] is the discrete Doppler spread vector of the DT domain channel, is the set of different time delays of P propagation paths in the DD domain, 0 ≤ l ≤ l max ;

[0072] is the transmitted signal in the DT domain The estimated value of the column vector of the symbol in the m-th row, and the value of l should ensure that m - l ≥ 0;

[0073] denotes the Hadamard product;

[0074] (5.2) Calculate the maximum ratio combining vector of the RNPI combination including the transmitted symbol vector x m in the DD domain at different time delays l

[0075]

[0076] where, (·) * denotes the complex conjugate matrix, and the value of l should ensure that m + l ≤ M - 1;

[0077] (5.3) Calculate the soft value estimation vector of the maximum ratio combination

[0078]

[0079] Among them, represents Hadamard division, (·) H is the conjugate transpose of the matrix, and the value of l should ensure that m + l ≤ M - 1;

[0080] (5.4) Converts the soft value estimation vector in the DT domain to the soft value estimation vector in the DD domain

[0081]

[0082] (5.5) Uses the LLR converter to convert it into bit-level LLR information

[0083]

[0084] Among them, is the LLR of the b-th bit of the n-th transmitted symbol in the soft value estimation vector in the DD domain where n = 0,..., N - 1, N is the total number of symbol blocks, b = 0, 1,..., log2(Q), Q is the modulation order, the value of the b-th bit is 0 or 1, Q0 and Q1 are subsets of the QAM symbol set, Q0 is the set of QAM symbols where the value of the b-th bit is 0, and Q1 is the set of QAM symbols where the value of the b-th bit is 1, is the variance of the normalized MRC posterior noise plus interference NPI vector;

[0085] Subsequently, is deinterleaved and LDPC decoded to generate hard decision encoded bits

[0086] (5.6) Interleaves and Q-order QAM modulates the hard decision encoded bits obtained by decoding again to obtain the reconstructed modulation vector By setting the weight factor δ, the soft value estimation vector of MRC is weighted with the reconstructed modulation vector obtained after decoding to obtain the estimated value of the DT domain transmitted symbol vector for the i-th iteration

[0087]

[0088]

[0088] (5.7) At different row indices m, and ensuring that when the value of l satisfies m + l ≤ M - 1, use the estimated value of the transmitted symbol vector in the DT domain of the i-th iteration Update the RNPI vector

[0089]

[0090] (5.8) The norm of the RNPI vector of the i-th iteration Compare with the norm of the RNPI vector of the previous iteration ;

[0091] If Then stop the iteration, and use the estimated value of the transmitted symbol vector in the DT domain of the i-th iteration As the final estimated value of the transmitted symbol vector in the DT domain For output;

[0092] Otherwise, execute step (5.9);

[0093] (5.9) Compare the current iteration number i with the maximum iteration number ite:

[0094] If i < ite, then set i = i + 1, and return to step (5.2);

[0095] Otherwise, use the estimated value of the transmitted symbol vector in the DT domain of the ite-th iteration As the final estimated value of the transmitted symbol vector in the DT domain For output;

[0096] Step 6, perform an N-point FFT transformation on the estimated value of the transmitted symbol vector in the DT domain To obtain the estimated value of the transmitted symbol vector in the DD domain

[0097]

[0098] For the estimated value of the transmitted symbol vector in the DD domain Perform LLR conversion, deinterleaving, and LDPC decoding decision in sequence to obtain the estimated value of the transmitted message bit sequence

[0099] Embodiment 2, an MRC signal detection device based on a weighted update mechanism.

[0100] Refer to Figure 3 , the device described in this example includes: an MRC detector module 1, an LLR converter module 2, a deinterleaving module 3, an LDPC decoder module 4, an interleaving module 5, a QAM modulation module 6, and a weighted update module 7, where:

[0101] The MRC detector module 1 is used to extract and combine the received multipath components of the transmitted symbols based on maximum ratio combining in the DT domain, improve the signal-to-noise ratio of the combined signal, enhance the reliability of the signal, and input the soft value information output by the MRC detector to the LLR converter module 2 to provide a more accurate input for the subsequent bit-level log-likelihood ratio (LLR) calculation;

[0102] The LLR converter module 2 is used to convert the soft value information output by the MRC detector into bit-level log-likelihood ratio (LLR) and input the bit-level LLR information to the deinterleaver module 3;

[0103] The deinterleaver module 3 is used to restore the bit-level LLR information rearranged by the interleaver to the original order to obtain the deinterleaved data, and input the deinterleaved data to the LDPC decoder module 4;

[0104] The LDPC decoder module 4 is used to perform LDPC decoding operation on the deinterleaved data to obtain hard-decision encoded bits, and input the hard-decision encoded bits to the interleaver module 5;

[0105] The interleaver module 5 is used to perform interleaving processing on the hard-decision encoded bits output by the LDPC decoder to obtain the interleaved data, and input the interleaved data to the QAM modulation module 6 to improve the anti-interference ability of the system against burst errors by dispersing consecutive errors occurring during the transmission process;

[0106] The QAM modulation module 6 is used to perform QAM modulation on the interleaved data again to obtain a reconstructed modulation vector containing decoding information, and input the reconstructed modulation vector to the weighted update module 7;

[0107] The weighted update module 7 is used to perform weighted combination of the soft value information output by the MRC detector and the reconstructed modulation signal output by the QAM modulation, optimize the overall bit error rate performance of the system by enhancing the interaction information between detection and decoding, and achieve more efficient signal detection and decoding.

[0108] The effects of the present invention can be further illustrated by the following simulations:

[0109] I. Simulation conditions:

[0110] The OTFS signal DD domain grid size adopted in the simulation is M×N = 64×16, and the carrier frequency f c= 4 GHz, subcarrier spacing Δf = 15 KHz; To simulate a high-speed mobile scenario, the user's moving speed is set to 500 km / h, and the corresponding maximum Doppler shift is 1.85 KHz. LDPC code is used for channel coding, the code rate R is set to 1 / 2, and the codeword length is 672 bits. In addition, the standard EVA channel model is used in the simulation to simulate the actual wireless propagation environment.

[0111] To ensure the statistical reliability of the simulation results, simulations of transmitting 5000 OTFS signals are carried out at each signal-to-noise ratio SNR point, and the maximum number of Turbo iterations ite is set to 20 times. Based on these data, the bit error rate BER curve is plotted.

[0112] II. Simulation Content and Result Analysis:

[0113] Simulation 1, under the 4-QAM modulation mode, the bit error rate curves with different weight factors δ are simulated using the present invention, where δ = 1 is the existing MRC detection algorithm applied to the coded OTFS system. The results are as Figure 4 shown.

[0114] From Figure 4 it can be seen that: when δ = 1, the system performance is the best. This is because in the low-order modulation scheme, the sparse constellation distribution ensures the reliability of QAM re-modulation during the Turbo iteration process. Therefore, it shows that for the 4-QAM modulation mode, the existing MRC detection selects δ = 1, that is, it completely depends on the LDPC decoding result to obtain better bit error rate performance.

[0115] Simulation 2, under the 16-QAM modulation mode, the bit error rate curves with different weight factors δ are simulated using the present invention, where δ = 1 is the existing MRC detection algorithm applied to the coded OTFS system. The results are as Figure 5 shown.

[0116] From Figure 5 it can be seen that: when δ = 0.5, the system performance is the best. Specifically, when BER = 1×10 -3 , when the present invention selects δ = 0.5, there is a 0.3 dB performance improvement compared with the existing MRC detection algorithm;

[0117] Simulation 3, under the 64-QAM modulation mode, the bit error rate curves with different weight factors δ are simulated using the present invention, where δ = 1 is the existing MRC detection algorithm applied to the coded OTFS system. The results are as Figure 6 shown.

[0118] From Figure 6 it can be seen that: when δ = 0.5, the system performance is the best. Specifically, when BER = 1×10 -3When δ = 0.5 is selected in the present invention, there is a 2.8 dB performance improvement compared with the existing MRC detection algorithm;

[0119] The simulation results show that in the case of low-order modulation, the sparse constellation distribution can effectively ensure the reliability of QAM re-modulation during the Turbo iteration process, enabling the existing MRC detection algorithm to fully utilize the reliability of LDPC decoding information, thereby achieving the best performance. However, in the case of high-order modulation scenarios, the performance of the existing MRC is significantly insufficient. This is mainly because the constellation points are densely distributed in the high-order modulation scheme, and the signal is easily interfered by noise or errors during the QAM re-modulation process, resulting in a decline in detection performance. In contrast, when δ = 0.5 is selected using the method of the present invention, it can not only use the reliable information after decoding to correct the result, but also not completely abandon the MRC soft value estimate of this iteration, balancing the result of the MRC soft value output and the LDPC decoding result to achieve the best performance.

Claims

1. A method for MRC signal detection based on a weighted update mechanism, characterized in that, Including: (1) At the transmitting end, the binary transmission message bit sequence u is encoded by a low-density parity-check (LDPC) code, interleaved, modulated, and mapped to generate a delay-Doppler (DD) domain transmission signal X. The DD domain transmission signal X is converted to obtain a time-domain transmission signal z, which is then transmitted through a wireless channel. (2) At the receiving end, the time-domain received signal r is matrix-reconstructed to obtain the delay-time DT-domain received signal (3) Using the Turbo iterative structure, perform MRC iterative detection based on the weighted update mechanism on the received signal in the DT domain that is, MRC detector, LLR converter, deinterleaving, LDPC decoder, interleaving, QAM modulation, and weighted update to obtain the estimated value of the transmitted symbol vector in the DT domain where m is the row index of the DD domain grid, δ is the weight factor, and 0 < δ ≤ 1; is the DT domain MRC soft value estimation vector output by the MRC detector, is the reconstructed modulation vector output after QAM modulation; (4) Estimate the transmitted symbol vector in the DT domain Convert it to the estimated transmitted symbol vector in the DD domain And then Perform LLR conversion, deinterleaving, and LDPC decoding decision in sequence to obtain the estimated value of the transmitted message bit sequence 2. The method according to claim 1, wherein In step (1), at the transmitting end, after the binary transmission message bit sequence u is encoded by a low-density parity-check (LDPC) code, interleaved, modulated, and mapped, a delay-Doppler (DD) domain transmission signal X is generated. The implementation includes the following: Set up a time-delay Doppler (DD) domain grid with dimensions of M×N, and set the symbols in the last M' rows of the grid to zero, where N is the total number of symbol blocks, M is the total number of subcarriers, and M'≥l max , l max is the maximum delay tap of the channel; At the transmitting end, the binary message bit sequence u ∈ (0, 1) k is encoded by an LDPC encoder with a code rate of R to obtain the encoded sequence c ∈ (0, 1) L , and then the encoded sequence c is passed through an interleaver to obtain the interleaved sequence c′, where k is the length of the message sequence and L is the total length after encoding; After performing Q-order quadrature amplitude modulation (QAM) on the interleaved sequence c′, it is then mapped into grid information symbols of (M - M′) × N in the DD domain to obtain the two-dimensional DD domain transmitted signal X, where represents the M × N dimensional matrix space.

3. The method according to claim 1, characterized in that, In step (1), the conversion of the DD domain transmission signal X to obtain the time-domain transmission signal z includes the following: First, perform an N-point inverse fast Fourier transform (N-IFFT) on the DD domain transmission signal X to obtain a discrete-time (DT) domain transmission signal X: Among them, represents the M×N dimensional matrix space, F N is the normalized N-point discrete Fourier transform DFT matrix, (·) H is the conjugate transpose of the matrix; Secondly, perform matrix column vectorization on the signal transmitted in the DT domain to obtain the time-domain transmitted signal z: Among them, represents the MN×1 dimensional matrix space, and vec(·) represents the column vectorization of the matrix.

4. The method according to claim 1, wherein The time-domain received signal r received at the receiving end in step (2) is expressed as follows: r = Hz + w Among them, w is a time-domain noise vector, represents the MN×1 dimensional matrix space, Among them represents the MN×MN dimensional matrix space, P is the number of propagation paths, and h q is the channel gain of the q-th path, and l q and k q are the delay component and Doppler frequency shift of the q-th path respectively; is a permutation matrix, used to simulate the effect of the delay of the q-th path; Δ = diag{α 0 , α 1 , α 2 ,..., α MN-1} is a diagonal matrix, which is used to simulate the influence of the Doppler frequency shift of the q-th path, α = e j2π / MN , j = 0,..., MN - 1; The DT-domain received signal obtained in the step (2) which is expressed as follows: where represents the M×N dimensional matrix space, is the inverse operation of vec(·), that is, reconstructing the vector into a matrix with M rows and N columns.

5. The method according to claim 1, characterized in that In step (3), the Turbo iterative structure is used to perform MRC iterative detection on the received signal in the DT domain based on a weighted update mechanism, and its implementation includes the following: (3a) Initialize the estimated value of the DT-domain transmitted symbol vector at the beginning of the iteration Set the maximum number of iterations ite, and calculate the initial residual noise and interference RNPI vector where, 0 N is a zero vector of length N, is the DT-domain transmitted signal the estimated value of the column vector of the m-th row symbol, m = 0,..., M-1, M is the total number of subcarriers, i is the iteration number, and i = 0 at the beginning of the iteration; (3b) Calculate the maximum ratio combining vector of the RNPI combination including the DD-domain transmitted symbol vector x m and the soft value estimation vector of the DT domain with maximum ratio combining where is a set of different delay amounts of P propagation paths in the DD domain, 0 ≤ l ≤ l max , l max is the maximum delay tap of the channel;​ (3c) Convert the soft value estimation vector in the DT domain to the soft value estimation vector in the DD domain Then, use the LLR converter to convert it to bit-level LLR information and perform deinterleaving and LDPC decoding on it to generate hard decision coded bits (3d) The hard-decision coded bits obtained by decoding are interleaved and Q-ary QAM modulated again to obtain a reconstructed modulation vector By setting a weight factor δ, the soft value estimation vector of MRC is weighted with the reconstructed modulation vector obtained after decoding to obtain an estimated value of the transmitted symbol vector in the DT domain for the i-th iteration (3e) At different row indices m, and ensuring that when the value of l satisfies m + l ≤ M - 1, use the estimated value of the transmitted symbol vector in the DT domain of the i-th iteration Update the RNPI vector (3f) Compare the RNPI vector norm of the $i$-th iteration with the RNPI vector norm of the previous iteration ; If stop the iteration and use the estimated symbol vector of the DT domain in the i-th iteration as the final estimated symbol vector of the DT domain for output; Otherwise, execute step (3g); (3g) Compare the current iteration number i with the maximum iteration number ite: If i < ite, then set i = i + 1 and return to step (3b); Otherwise, send the estimated symbol vector of the DT domain in the ite-th iteration as the estimated symbol vector of the final DT domain to be sent for output.

6. The method according to claim 5, characterized in that Calculate the initial residual noise and interference RNPI vector in step (3a). The formula is as follows: where i is the iteration number, and at the start of the iteration, i = 0; Estimate the signal for the DT domain The column vector of the symbols in the m-th row, where m = 0, ..., M-1 and M is the total number of subcarriers; is the discrete Doppler spread vector of the DT domain channel, is the set of different time delay amounts of P propagation paths in the DD domain, 0 ≤ l ≤ l max , l max is the maximum delay tap of the channel; Transmit signal for the DT domain The estimated value of the column vector of the symbols in the m-th row, where the value of l should ensure that m - l ≥ 0; Denotes the Hadamard product.

7. The method according to claim 5, wherein Calculate the maximum ratio combining vector of the RNPI combinations including the DD domain transmitted symbol vector x m in step (3b), and the soft value estimation vector of the DT domain with maximum ratio combining The formulas are as follows: Among them, is the discrete Doppler spread vector of the DT domain channel. The value of l should ensure that m + l ≤ M - 1, and (·) * represents the complex conjugate matrix, represents the Hadamard product, where m = 0,..., M - 1 and M is the total number of subcarriers; Denotes Hadamard division, (·) H Is the conjugate transpose of a matrix.

8. The method according to claim 5, wherein In the step (3c), the soft value estimation vector in the DT domain is converted into the soft value estimation vector in the DD domain The formula is as follows: Among them, F N is a normalized N-point discrete Fourier transform DFT matrix; In the step (3c), the LLR converter is used to convert it into bit-level LLR information The formula is as follows: Among them, is the soft value estimation vector of the DD domain is the LLR of the b-th bit of the n-th transmitted symbol in, where n = 0,..., N-1, N is the total number of symbol blocks, b = 0, 1,..., log2(Q), Q is the modulation order, the value of the b-th bit is 0 or 1, Q0 and Q1 are subsets of the QAM symbol set, Q0 is the set of values of the b-th bit in the QAM symbol set being 0, and Q1 is the set of values of the b-th bit in the QAM symbol set being 1. is the variance of the normalized MRC posterior noise plus interference NPI vector.

9. The method according to claim 5, characterized in that, Updating the RNPI vector in step (3e) The formula is as follows: Among them, is the discrete Doppler spread vector of the DT domain channel, represents the Hadamard product. The value of l should ensure that m + l ≤ M - 1, where M is the total number of subcarriers.

10. An MRC signal detection device based on a weighted update mechanism, characterized in that, Including: An MRC detector module for extracting and combining the received multipath components of the transmitted symbols based on maximum ratio combining in the DT domain to improve the signal-to-noise ratio of the combined signal, enhance the reliability of the signal, and provide more accurate input for subsequent bit-level log-likelihood ratio (LLR) calculation; An LLR converter module for converting the soft value information output by the MRC detector into bit-level log-likelihood ratio (LLR) to provide reliability information for subsequent channel decoding; An interleaving module for restoring the bit-level LLR information rearranged by the interleaver to its original order; An LDPC decoder module for performing LDPC decoding operations on the data after deinterleaving to obtain hard-decision coded bits; An interleaving module for interleaving the hard-decision coded bits output by the LDPC decoder, that is, by rearranging the order of the data, so that consecutive errors occurring during transmission are dispersed, enhancing the system's resistance to burst errors; A QAM modulation module for performing QAM modulation on the interleaved data again to obtain a reconstructed modulation vector containing decoding information; A weighted update module for weighted combining the soft value information output by the MRC detector and the reconstructed modulation signal output by the QAM modulation to effectively enhance the interaction information between detection and decoding and achieve better bit error rate performance.

Images