An iterative decoding feedback detection method for underwater acoustic OTFS communication system

Abstract

The present application discloses an iterative decoding feedback detection method for an underwater acoustic OTFS communication system, which relates to the field of information technology. The method comprises: constructing an outer loop iteration structure between a delay-Doppler domain equalizer and an LDPC decoder, utilizing the delay-Doppler domain equalizer in the underwater acoustic OTFS communication system to output symbol-level external information based on equalizer prior information and generate decoding prior information. Utilizing the LDPC decoder to decode and output bit-level LLR information based on the decoding prior information, and generate equalizer prior information. Repeating the outer loop iteration process until the number of outer loop iterations reaches a maximum value, performing soft decision decoding based on the bit-level LLR information output by the LDPC decoder, and outputting a bit stream detection result. Iteratively interacting the symbol-level external information output by the equalizer and the bit-level LLR information output by the decoder for multiple times can reduce the loss of useful information and improve system performance.

Description

Technical Field

[0001] The present application relates to the field of information technology, and in particular to an iterative decoding feedback detection method for an underwater acoustic (OTFS) communication system. Background Art

[0002] OFDM (Orthogonal Frequency Division Multiplex) is a multi-carrier modulation technology widely used in traditional wireless communication systems, greatly improving the spectrum efficiency of wireless communication systems. However, for underwater acoustic channels, it has the characteristics of limited bandwidth, large delay spread, frequency-selective fading, and susceptibility to the Doppler effect. In high-speed mobile communication scenarios such as underwater acoustic channels, the Doppler frequency offset caused by the relative motion of the system transmitter and receiver will destroy the orthogonality between OFDM subcarriers, thereby generating inter-carrier interference and significantly reducing OFDM performance. Some methods also try to reduce the channel variation of each symbol by shortening the OFDM symbol duration, but this approach will reduce the spectrum efficiency due to the shortened symbol duration and cyclic prefix. These problems have led to unsatisfactory application of OFDM in underwater acoustic channels.

[0003] OTFS (Orthogonal Time Frequency Space) technology can compensate for the significant performance degradation of OFDM in high-speed mobile communication scenarios. It is a multi-carrier modulation technique used to improve and address the severe Doppler effect and channel effects in high-speed mobile communication systems. OTFS modulates information in the delayed Doppler domain, offering strong delay and Doppler tolerance and the potential for full diversity, making it key to supporting reliable communications. Underwater OTFS communication systems improve the frequency efficiency of underwater acoustic communication networks in high-Doppler, high-latency underwater acoustic channels, and have become a major research area in underwater acoustic communication technology in recent years.

[0004] The system structure diagram of the common classic underwater acoustic OTFS communication system is as follows: Figure 1As shown, at the transmitting end of the underwater acoustic OTFS communication system, the transmitted information bit stream is encoded by the LDPC encoder and then modulated into OTFS symbols by the symbol modulator and placed in the delay-Doppler domain grid. After the inverse symplectic finite fourier transform (ISFFT) and Heisenberg transform, it is converted into a time domain signal, and finally reaches the receiving end after passing through the underwater acoustic time-varying channel. At the receiving end of the underwater acoustic OTFS communication system, the received time domain signal is converted into each domain after the symplectic finite fourier transform (SFFT) and Wigner transform, and then the information is detected by the equalizer of each domain. The information output by the equalizer is input into the LDPC decoder for decoding. The LDPC decoder decodes the received information bits and outputs them. Figure 1 It can be seen that the current receiving end of the underwater acoustic OTFS communication system mainly adopts a cascade structure of an equalizer and an LDPC decoder. However, since the received time domain signal usually contains a large amount of noise signals, this cascade structure has a high bit error rate under the influence of noise, which affects the reliability and system performance of the underwater acoustic OTFS communication system. Summary of the Invention

[0005] In response to the above-mentioned problems and technical requirements, this application proposes an iterative decoding feedback detection method for an underwater acoustic OTFS communication system. The technical solution of this application is as follows:

[0006] An iterative decoding feedback detection method for an underwater acoustic OTFS communication system comprises the following steps:

[0007] The iterative decoding feedback detection method includes out In the outer loop iteration:

[0008] In n out = 1, the delay-Doppler domain equalizer in the underwater acoustic OTFS communication system is used to perform multiple time-domain Doppler detections on the acquired delay-Doppler domain signal to be detected and output the symbol-level external information; out ≥2, using the delay-Doppler domain equalizer according to the nth out The equalizer prior information obtained by -1 outer loop iteration outputs symbol-level external information; the symbol-level external information indicates any OTFS symbol x in the delay-Doppler domain signal to be detected. i The delay-Doppler domain equalizer detects the modulation symbol a in the modulation symbol information set A. j Probability a j ∈Α;

[0009] Generate decoding prior information based on symbol-level extrinsic information Decoding prior information Indicates OTFS symbol x i The kth bit c of i,k The log-likelihood ratio of

[0010] The LDPC decoder in the underwater acoustic OTFS communication system is used to decode the prior information Perform multiple LDPC decoding and output bit-level LLR information

[0011] When n out <n out_max When, according to Generate nth out The equalizer prior information of the outer loop iteration enters the next outer loop iteration, and the equalizer prior information indicates any OTFS symbol x in the delay-Doppler domain signal to be detected i Detected by the LDPC decoder as modulation symbol a j Probability

[0012] When n out =n out_max When, according to Perform soft decision decoding and output the bit stream detection result of the delay Doppler domain signal to be detected. The integer parameter n out_max ≥2.

[0013] Its further technical solution is to generate decoding prior information based on symbol-level external information include:

[0014] The LLR converter in the underwater acoustic OTFS communication system is used to calculate the OTFS symbol x in the interleaved state according to the principle of maximum a posteriori probability criterion. i The kth bit c of i,k The log-likelihood ratio

[0015] The deinterleaver in the underwater acoustic OTFS communication system is used to deinterleave the OTFS symbol x in the interleaved state. i The kth bit c of i,k The log-likelihood ratio Perform deinterleaving to obtain decoding prior information

[0016] Its further technical solution is, according to Generating updated equalizer prior information includes:

[0017] Using the interleaver in underwater acoustic OTFS communication system to align bit-level LLR information Perform interleaving processing to obtain the OTFS symbol x in the interleaved statei The kth bit c detected i,k The log-likelihood ratio

[0018] The LLR converter in the underwater acoustic OTFS communication system is used to calculate the OTFS symbol x in the interleaved state. i The kth bit c of i,k The log-likelihood ratio The updated equalizer prior information is obtained by conversion.

[0019] Its further technical solution is that the OTFS symbol x in the interleaved state i The kth bit c of i,k The log-likelihood ratio is the set of modulation symbols whose k-th bit is 1 in the modulation symbol information set A, It is the set of modulation symbols whose k-th bit is 0 in the modulation symbol information set A.

[0020] Its further technical solution is, according to Generate nth out The equalizer prior information for the outer loop iteration includes:

[0021] According to the OTFS symbol x in the interleaved state i The kth bit c detected i,k The log-likelihood ratio Calculate the bit c separately i,k The prior probability p(c i,k =0) and bit c i,k The prior probability p(c i,k =1);

[0022] Use LLR converter to calculate the bit position c i,k The prior probability p(c i,k =0) and the prior probability p(c i,k =1) Convert to get the nth out The equalizer prior information for the outer loop iteration.

[0023] Its further technical solution is that bit c i,k The prior probability p(c i,k )for:

[0024]

[0025] Among them, when c i,k = 0, λ m,k =-1, When c i,k =1, λm,k =1,

[0026] Its further technical solution is to convert the nth out The equalizer prior information for the outer loop iteration includes determining the OTFS symbol x i Detected by the LDPC decoder as modulation symbol a j Probability for:

[0027]

[0028] The modulation symbol information set A includes Q modulation symbols. For any j∈[1,Q], the modulation symbol a j is a binary sequence of length log2|Q|, p(c i,k =a j,k ) is to convert the OTFS symbol x i The kth bit c detected i,k As the modulation symbol a j The kth bit a j,k When bit a j,k The prior probability of , Π is the multiplication operator symbol.

[0029] A further technical solution is that the iterative decoding feedback detection method further includes:

[0030] Channel estimation is performed on the time-domain received signal r(t) obtained through the underwater acoustic time-varying channel to obtain a channel equivalent matrix, and sigmoid Fourier transform and Wigner transform are performed on the time-domain received signal r(t) to obtain the time-delayed Doppler signal to be detected;

[0031] In n out = 1, a delay-Doppler domain equalizer is used to perform multiple time-domain Doppler detections based on the channel equivalent matrix and the delay-Doppler domain signal to be detected, and output symbol-level external information;

[0032] In n out ≥2, using the delay-Doppler domain equalizer according to the channel equivalent matrix and the nth out The equalizer prior information obtained by -1 outer loop iteration outputs symbol-level external information.

[0033] The beneficial technical effects of this application are:

[0034] This application proposes an iterative decoding feedback detection method for an underwater acoustic (OTFS) communication system. This method incorporates an external loop iteration structure into the cascaded structure of the system's delay-Doppler domain equalizer and LDPC decoder. By iteratively interacting the symbol-level external information output by the delay-Doppler domain equalizer with the bit-level LLR information output by the LDPC decoder, the method reduces the loss of useful information, thereby lowering the system's bit error rate and improving the performance of the underwater acoustic (OTFS) communication system.

[0035] By using the bit-level LLR information of the LDPC decoder for soft-decision decoding, the decoding results can be brought close to the Shannon limit, achieving highly reliable decoding results. For low signal-to-noise ratios during multiple iterations, soft decisions take into account the noise in the received signal and estimate the probability of the received signal, resulting in better decoding performance in high-noise conditions. The outer loop iteration process also improves the accuracy of the decision results. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 This is the structure diagram of the classic underwater acoustic OTFS communication system.

[0037] Figure 2 It is a structural diagram of the iterative decoding feedback detection system.

[0038] Figure 3 Flowchart of iterative decoding feedback detection method

[0039] Figure 4 This is a comparison chart of bit error rates for different numbers of outer loop iterations. DETAILED DESCRIPTION

[0040] The specific implementation of this application will be further described below with reference to the accompanying drawings.

[0041] This application discloses an iterative decoding feedback detection method for underwater acoustic OTFS communication system. Figure 2 The iterative decoding feedback detection system shown in FIG. Figure 1 Based on the cascaded delay-Doppler domain equalizer and LDPC decoder in the classic underwater acoustic OTFS communication system shown in the figure, an information processing device is added to the path from the delay-Doppler domain equalizer to the LDPC decoder, and a feedback connection from the decoder to the delay-Doppler domain equalizer is added, so that external information exchange can be carried out between the delay-Doppler domain equalizer and the LDPC decoder, thereby fully utilizing useful information.

[0042] Please refer to Figure 3 As shown in the flowchart, the iterative decoding feedback detection method is used at any nth out In the outer loop iteration:

[0043] Step 1, in n out = 1, the delay-Doppler domain equalizer in the underwater acoustic OTFS communication system performs multiple time-domain Doppler detections on the acquired delay-Doppler domain signal to be detected and outputs symbol-level external information. The delay-Doppler domain equalizer calculates the symbol-level external information by internally performing multiple time-domain Doppler detections. The specific number of executions can be set according to the detection effect. The delay-Doppler domain equalizer is used in n out ≥2, using the delay-Doppler domain equalizer according to the nth out The equalizer prior information obtained by -1 outer loop iterations outputs symbol-level external information. The symbol-level external information indicates any OTFS symbol x in the delay-Doppler domain signal to be detected. i The delay-Doppler domain equalizer detects the modulation symbol a in the modulation symbol information set A. j Probability a j ∈Α. The modulation symbol information is generated by digital signal modulation of the transmitted information bit stream of the underwater acoustic OTFS communication system through a symbol modulator. The digital signal modulation method can adopt QAM (Quadrature Amplitude Modulation). QAM maps the input bits to the constellation diagram to form a modulation symbol. Each constellation point in the constellation diagram corresponds to a modulation symbol. The transmitted information bit stream is subjected to QAM to obtain the modulation symbol information set A. The number of modulation symbols contained in A is Q, and the corresponding modulation mode is Q-QAM. The value of Q can be set according to actual needs. Generally, 16, 64, and 256 can be selected, corresponding to different modulation symbol forms. For example, when the modulation mode is 16-QAM, the modulation symbol information set A contains 16 modulation symbols, and each modulation symbol a j is a binary sequence of length 4.

[0044] In one embodiment, an iterative decoding feedback detection system performs channel estimation on the acquired time-domain received signal r(t) through an underwater acoustic time-varying channel to obtain a channel equivalent matrix H. The time-domain received signal r(t) is then subjected to a sigmoid Fourier transform and a Wigner transform to obtain the delay-Doppler signal r(τ,υ) to be detected. Channel estimation, sigmoid Fourier transform, and Wigner transform are existing technologies used and are not further described in this application.

[0045] In n out = 1, the delay-Doppler domain equalizer performs multiple time-domain Doppler detections based on the channel equivalent matrix and the delay-Doppler domain signal to be detected and outputs symbol-level external information. out ≥2, using the delay-Doppler domain equalizer according to the channel equivalent matrix and the nth out The equalizer prior information obtained by -1 outer loop iteration outputs symbol-level external information.

[0046] Step 2: Generate decoding prior information based on symbol-level external information Decoding prior information Indicates OTFS symbol x i The kth bit c of i,k The log-likelihood ratio.

[0047] When transmitting underwater acoustic data, the underwater acoustic time-varying channel needs to meet the requirements of real-time and high speed. This requires the underwater acoustic OTFS communication system to fully consider the effects of decoding effect, computational complexity, and system overhead in its design, ensuring the simplification of signal processing devices on the system path to achieve accurate and efficient transmission of underwater acoustic data. By directly transmitting the symbol-level external information of the received signal of the detection system in the delay-Doppler domain, the effective transmission of the received signal is guaranteed, unnecessary information loss is reduced, and the efficiency and accuracy of signal transmission are improved. In one embodiment, the LLR converter in the underwater acoustic OTFS communication system is used to calculate the OTFS symbol x in the interleaved state based on the symbol-level external information output by the delay-Doppler domain equalizer according to the principle of the maximum a posteriori probability criterion. i The kth bit c of i,k The log-likelihood ratio The deinterleaver in the underwater acoustic OTFS communication system is used to deinterleave the OTFS symbol x in the interleaved state. i The kth bit c of i,k The log-likelihood ratio Perform deinterleaving to obtain decoding prior information Input to the LDPC decoder.

[0048] In one embodiment, the OTFS symbol x in the interleaved state i The kth bit c of i,k The log-likelihood ratio is the set of modulation symbols whose k-th bit is 1 in the modulation symbol information set A, It is the set of modulation symbols whose k-th bit is 0 in the modulation symbol information set A.

[0049] Step 3: using the LDPC decoder in the underwater acoustic OTFS communication system to decode the prior information Perform multiple LDPC decoding and output bit-level LLR information The LDPC decoder calculates the bit-level LLR information by performing multiple decoding processes internally. The specific number of executions can be set according to the decoding effect.

[0050] Step 4, when n out <n out_max When, according to Generate nthout The equalizer prior information of the outer loop iteration enters the next outer loop iteration, and the equalizer prior information indicates any OTFS symbol x in the delay-Doppler domain signal to be detected i Detected by the LDPC decoder as modulation symbol a j Probability

[0051] In one embodiment, the interleaver in the underwater acoustic OTFS communication system is used to interleave the bit-level LLR information. Perform interleaving processing to obtain the OTFS symbol x in the interleaved state i The kth bit c detected i,k The log-likelihood ratio The LLR converter in the underwater acoustic OTFS communication system is used to calculate the OTFS symbol x in the interleaved state. i The kth bit c of i,k The log-likelihood ratio The updated equalizer prior information is obtained by conversion.

[0052] Step 5, when n out =n out_max When, according to Perform soft decision decoding and output the bit stream detection result of the delay Doppler domain signal to be detected. The integer parameter n out_max ≥2.n out_max It is the maximum number of iterations of the outer loop, which can be customized according to the detection effect. i,k The log-likelihood ratio The size of c is determined by i,k In one embodiment, when When it is greater than 0, determine c i,k The bit value of c is 0; otherwise, i,k The bit value is 1. According to each bit c i,k The bit values ​​of the output bit stream detection result can be composed.

[0053] In another embodiment, in step 4, Generate nth out The specific method of the equalizer prior information of the outer loop iteration is as follows: according to the OTFS symbol x in the interleaved state i The kth bit c detected i,k The log-likelihood ratio Calculate the bit c separately i,k The prior probability p(c i,k =0) and bit c i,k The prior probability p(c i,k=1). LLR converter is used to calculate the bit position c. i,k The prior probability p(c i,k =0) and the prior probability p(c i,k =1) Convert to get the nth out The equalizer prior information for the outer loop iteration.

[0054] In one embodiment, bit c i,k The prior probability p(c i,k )for:

[0055]

[0056] Among them, when c i,k = 0, λ m,k =-1, When c i,k =1, λ m,k =1,

[0057] In one embodiment, the OTFS symbol x is determined i Detected by the LDPC decoder as modulation symbol a j Probability for:

[0058]

[0059] The modulation symbol information set A includes Q modulation symbols. For any j∈[1,Q], the modulation symbol a j is a binary sequence of length log2|Q|, p(c i,k =a j,k ) is to convert the OTFS symbol x i The kth bit c detected i,k As the modulation symbol a j The kth bit a j,k When bit a j,k The prior probability of , Π is the multiplication operator symbol. Modulation symbol a j With log2|Q| bits, the modulation symbol a can be obtained by multiplying the prior probability of each bit j The prior probability of , and then get the OTFS symbol x i Detected by the LDPC decoder as modulation symbol a j probability.

[0060] The iterative decoding feedback detection method proposed in this application is used to calculate the bit error rate under different outer loop iteration times. Figure 4As shown in the figure, the bit error rate of multiple information interactions by setting outer loop iterations is lower than that without setting outer loop iterations, and the bit error rate decreases with the increase of the number of outer loop iterations, which can effectively improve the performance of the underwater acoustic OTFS communication system.

[0061] The above description is only a preferred embodiment of the present application, and the present application is not limited to the above embodiments. It is understood that other improvements and variations directly derived or imagined by those skilled in the art without departing from the spirit and concept of the present application should be considered to be included in the scope of protection of the present application.

Claims

1. An iterative decoding feedback detection method for an underwater acoustic OTFS communication system, characterized in that: The iterative decoding feedback detection method includes: out In the outer loop iteration: In n out = 1, the delay-Doppler domain equalizer in the underwater acoustic OTFS communication system is used to perform multiple time-domain Doppler detections on the acquired delay-Doppler domain signal to be detected and output the symbol-level external information; out ≥2, using the delay-Doppler domain equalizer according to the nth out The equalizer prior information obtained by -1 outer loop iteration outputs symbol-level external information; the symbol-level external information indicates any OTFS symbol x in the delay-Doppler domain signal to be detected. i The delay-Doppler domain equalizer detects the modulation symbol a in the modulation symbol information set A. j Probability a j ∈Α; Generate decoding priori information based on the symbol-level external information The decoding prior information Indicates OTFS symbol x i The kth bit c of i,k The log-likelihood ratio of The LDPC decoder in the underwater acoustic OTFS communication system is used according to the decoding prior information Perform multiple LDPC decoding and output bit-level LLR information When n out <n out_max When, according to Generate nth out The equalizer prior information of the outer loop iteration is obtained and the next outer loop iteration is entered, wherein the equalizer prior information indicates any OTFS symbol x in the delay-Doppler domain signal to be detected. i Detected by the LDPC decoder as modulation symbol a j Probability When n out =n out_max When, according to Perform soft decision decoding and output the bit stream detection result of the delay Doppler domain signal to be detected, the integer parameter n out_max ≥2.

2. The iterative decoding feedback detection method according to claim 1, wherein: Generating decoding prior information according to the symbol-level external information include: The LLR converter in the underwater acoustic OTFS communication system is used to calculate the OTFS symbol x in the interleaved state according to the principle of the maximum a posteriori probability criterion. i The kth bit c of i,k The log-likelihood ratio The deinterleaver in the underwater acoustic OTFS communication system is used to deinterleave the OTFS symbol x in the interleaved state. i The kth bit c of i,k The log-likelihood ratio Perform deinterleaving to obtain decoding prior information 3. The iterative decoding feedback detection method according to claim 1, wherein: The basis Generate nth out The equalizer prior information for the outer loop iteration includes: The interleaver in the underwater acoustic OTFS communication system is used to interleave the bit-level LLR information Perform interleaving processing to obtain the OTFS symbol x in the interleaved state i The kth bit c detected i,k The log-likelihood ratio The LLR converter in the underwater acoustic OTFS communication system is used according to the OTFS symbol x in the interleaved state. i The kth bit c of i,k The log-likelihood ratio Generate nth out The equalizer prior information for the outer loop iteration.

4. The iterative decoding feedback detection method according to claim 2, wherein: OTFS symbol x in interleaved state i The kth bit c of i,k The log-likelihood ratio is a set of modulation symbols whose k-th bit is 1 in the modulation symbol information set A, It is a set of modulation symbols whose k-th bit is 0 in the modulation symbol information set A.

5. The iterative decoding feedback detection method according to claim 3, wherein: according to Generate nth out The equalizer prior information for the outer loop iteration includes: According to the OTFS symbol x in the interleaved state i The kth bit c detected i,k The log-likelihood ratio Calculate the bit c separately i,k The prior probability p(c i,k =0) and bit c i,k The prior probability p(c i,k =1); Use LLR converter to calculate the bit position c i,k The prior probability p(c i,k =0) and the prior probability p(c i,k =1) Convert to get the nth out The equalizer prior information for the outer loop iteration.

6. The iterative decoding feedback detection method according to claim 5, wherein: Bit c i,k The prior probability p(c i,k )for: Among them, when c i,k = 0, λ m,k =-1, When c i,k =1, λ m,k =1, 7. The iterative decoding feedback detection method according to claim 5, wherein: Convert to get the nth out The equalizer prior information for the outer loop iteration includes determining the OTFS symbol x i Detected by the LDPC decoder as modulation symbol a j Probability for: The modulation symbol information set A includes Q modulation symbols. For any j∈[1,Q], the modulation symbol a j is a binary sequence of length log2|Q|, p(c i,k =a j,k ) is to convert the OTFS symbol x i The kth bit c detected i,k As the modulation symbol a j The kth bit a j,k When bit a j,k The prior probability of , Π is the multiplication operator symbol.

8. The iterative decoding feedback detection method according to claim 1, wherein: The iterative decoding feedback detection method further includes: Channel estimation is performed on the time-domain received signal r(t) obtained through the underwater acoustic time-varying channel to obtain a channel equivalent matrix, and sigmoid Fourier transform and Wigner transform are performed on the time-domain received signal r(t) to obtain the time-delayed Doppler signal to be detected; In n out =1, using a delay-Doppler domain equalizer to perform multiple time-domain Doppler detections based on the channel equivalent matrix and the delay-Doppler domain signal to be detected and output symbol-level external information; In n out ≥2, using the delay-Doppler domain equalizer according to the channel equivalent matrix and the nth out The equalizer prior information obtained by -1 outer loop iteration outputs symbol-level external information.

Images