Joint processing method of MPSK demodulation and Doppler tracking for underwater acoustic direct spread system

Abstract

The present invention discloses a joint processing method for MPSK demodulation and Doppler tracking for an underwater acoustic direct spread spectrum system, belonging to the field of underwater wireless communication and its signal processing. Specifically, the method comprises the following steps: first, multi-channel array elements of arbitrary formations simultaneously receive underwater direct sequence spread spectrum communication signals and perform preprocessing; then, each array element performs signal synchronization and Doppler coarse estimation; the received signal of each array element is de-carriered to obtain a baseband signal; then, the passband signal of each channel is despread to obtain a despread related signal and extract signal peak information; further, fine Doppler estimation and compensation are performed on each channel, and multi-channel soft value information is merged. Finally, the merged signal is decoded using BICM-ID to obtain the soft value information of each symbol under high-order modulation, and after decoding, the information bits carried by each symbol are obtained. The present invention realizes the joint integration technology of underwater communication and measurement and has certain anti-noise performance.

Description

Technical Field

[0001] The invention belongs to the field of underwater wireless communication and signal processing thereof, and particularly relates to a MPSK demodulation and Doppler tracking joint processing method for an underwater acoustic direct spread system. Background Art

[0002] Underwater direct-spectrum spread spectrum communication technology is widely used in underwater wireless communications and signal processing due to its good multipath resolution capability, strong spread spectrum gain, and support for multi-user simultaneous transmission. However, compared with single-carrier signals and multi-carrier signals, spread spectrum signals have a low transmission rate and long duration. In addition, due to the time-varying characteristics of the channel, the Doppler effect is particularly severe and presents a dynamic characteristic.

[0003] Traditional underwater Doppler mitigation methods can be summarized into two types: one is the differential mitigation method, and the other is the Doppler estimation and compensation method. Although the first differential mitigation method can directly resist Doppler interference, it will result in a loss of signal-to-noise ratio, and multiple differentials must be performed to resist time-varying Doppler, which leads to multiple amplification of noise. The traditional time-varying Doppler estimation method is highly complex and can only estimate the average speed, but cannot track the Doppler changes of each symbol in the direct-spread signal. The direct-spread signal has a long frame length, and the performance of the method using average speed for compensation is limited.

[0004] After literature search, it was found that the following literature has studied the Doppler tracking and estimation methods of underwater spread spectrum signals:

[0005] DJSun, et al. A symbol-based passband doppler tracking and compensation algorithm for underwater acoustic DSSS communications. Journal of Communications and Information Networks, 5(2):168–176, 2020. (hereinafter referred to as Document 1)

[0006] DJSun, et al. Iterative double-differential direct-sequence spreadspectrum reception in underwater acoustic channel with time-varying Dopplershifts. 22 The Journal of the Acoustical Society of America, 153(2):1027–1041, Feb. 2023 (hereinafter referred to as Document 2)

[0007] Reference 1 estimates the time-varying Doppler of the channel by using the time variation of the correlation peaks of two adjacent symbols. Although it can estimate the Doppler of each symbol more accurately, it is susceptible to multipath interference and has poor stability.

[0008] Reference 2 uses the phase change between adjacent symbols to estimate the time-varying Doppler of the channel. This method is not affected by multipath interference and has high accuracy. However, this method is affected by the phase ambiguity period and the Doppler dynamic range estimation is limited.

[0009] In addition, the methods in the above two documents are both affected by high-order modulation. When the signal modulates high-order information bits, the time delay or phase between adjacent symbols will change accordingly. Therefore, how to achieve stable time-varying Doppler estimation under the influence of high-order information bit modulation is an existing problem and one of the issues that urgently needs to be solved. Summary of the Invention

[0010] In response to the shortcomings of the existing technology, the present invention proposes a joint processing method of MPSK demodulation and Doppler tracking for underwater acoustic direct-spread spectrum systems. By adopting double differential modulation and demodulation technology, time-varying Doppler tracking, estimation and compensation under high-order modulation background, and iterative decoding, the method can achieve robust communication of high-order direct-spread spectrum signals in underwater acoustic time-varying channels and complete Doppler measurement of underwater acoustic time-varying channels.

[0011] The MPSK demodulation and Doppler tracking joint processing method for an underwater acoustic direct spread system comprises the following steps:

[0012] Step 1: Multi-channel array elements based on the ultra-short baseline array simultaneously receive the same underwater direct-spread communication signal and perform pre-processing;

[0013] Preprocessing includes bandpass filtering and downsampling;

[0014] The frame structure of underwater direct spread communication signals consists of a synchronization header signal and several groups of data. The synchronization header is modulated by a spread spectrum code, and each group of data is modulated by other spread spectrum codes of the same length.

[0015] Step 2: For the first array element, perform synchronization capture processing on the synchronization header of the received direct spread communication signal, and use a parallel filter bank to search the starting position and coarse Doppler velocity estimate of the array element received signal in the delay Doppler two-dimensional plane.

[0016] There are L array elements in total, with values ​​ranging from 0 to L-1;

[0017] Specifically:

[0018] Step 1-1, set the Doppler speed search interval Δυ and the maximum Doppler speed υ M , and then generate a parallel filter bank containing different Doppler velocities. The Doppler velocity vector υ of the filter bank is expressed as:

[0019] υ={-υ M ,-υ M +Δυ,…,0,…,υ M -Δυ,υ M}

[0020] Step 1-2: Generate n sets of local complex passband reference signals s with different pulse widths using the Doppler velocity vector υ n (t), the received signal of the lth array element is compared with the n-group reference signal s n (t) is matched and correlated, and the absolute value is taken to obtain the envelope of each branch, and then the delay-Doppler two-dimensional correlation matrix R(τ,υ) is obtained.

[0021] Step 1-3: Calculate the decision threshold based on the mean and variance of the signal noise segment and use it to detect whether the current signal contains the expected signal; if so, proceed to step 1-4; otherwise, update the signal and return to step 1-2.

[0022] Step 1-4: In the delay-Doppler two-dimensional plane, use the matrix R(τ,υ) to search for the highest peak value among the n groups of correlation results, and then obtain the delay τ corresponding to the received signal of the array element. l,0 , according to the time delay, the starting position of the lth array element receiving the spread spectrum signal is determined, and the desired communication signal r is intercepted according to the value l (t) and obtain the initial rough Doppler velocity estimate

[0023] Step 3: The communication signal r intercepted based on the starting position of the signal received by the lth array element l (t), perform down-conversion processing to obtain the baseband signal y l (t);

[0024] communication signalr l The expression of (t) is:

[0025]

[0026] where v k is the information bit of the kth symbol, A l,k is the signal amplitude of the kth symbol of the lth array element signal, υ l,k is the Doppler velocity value of the symbol, f c is the carrier center frequency of the signal, T s is the duration of a symbol, N d is the number of symbols in a frame signal, the operation symbol represents the real part, n(t) is the noise of the passband signal, and c1(t) is the despreading code of the local baseband information bit.

[0027] Down-conversion operation on communication signals l (t) is multiplied by the carrier and passed through a low-pass filter to obtain the baseband signal y l (t):

[0028]

[0029] Where LPF(·) represents low-pass filtering of the signal, ω l (t) represents the noise of the baseband received signal.

[0030] Step 4: Based on the rough Doppler velocity estimate of the lth array element signal and baseband signal y l (t) despreads the information bits of the spreading code c1(t) to generate the local baseband reference signal

[0031] Where c is the speed of sound in water;

[0032] Step 5: Based on local baseband reference signal Despread the correlation peak of each symbol in the received signal of the lth array element to obtain the despread correlation signal R l (τ);

[0033] Step 6: According to the correlation signal R of the lth array element l (τ), obtain the highest peak amplitude and phase of each symbol, and calculate the received signal corresponding to each symbol of the array element;

[0034] For the kth symbol of the array element, extract its highest peak amplitude and phase information Get the received symbol of the array element k∈{1,2,…,N d +2};

[0035] Step 7: The traditional PSP algorithm is improved by replacing the Viterbi algorithm with the BCJR algorithm. The Doppler shift is estimated and compensated on the branch metric corresponding to each symbol. The branch metric soft value information of the same symbol of all elements is combined to obtain the new soft value information γ k ;

[0036] The specific process is:

[0037] Step 1-1: According to the improved PSP algorithm, traverse the state quantity corresponding to each branch on the Trellis graph to obtain the symbol information bit corresponding to each state;

[0038] The information bit corresponding to the kth symbol state is b k ;

[0039] Step 1-2: For the kth symbol, use the information bits and the received signal y processed symbol by symbol l (k), estimate the Doppler phase offset for this symbol time for:

[0040]

[0041] Step 1-3: Compensate the Doppler phase offset of the k-th branch Get the metric soft value information γ of the branch l,k ;

[0042] Calculated as:

[0043]

[0044] where u k =[u k,1 ,…,u k,j ,…,u k,J ],u k,j ∈{1,-1} represents the information bits transmitted; and u k satisfy:

[0045]

[0046] where j = 1,…,J; b k is a double-difference coding sequence that satisfies

[0047] N is the number of multi-symbols used, L a (u k,j ) is the symbol u k,j The prior information of is 0 when it is first iterated, M is the modulation order of the multi-level phase shift, is the variance of the noise.

[0048] Step 1-4: Similarly, obtain the metric soft value information of each symbol branch of the l-th array element;

[0049] Step 1-5: Select the next array element and repeat the above steps to obtain the metric soft value information of each symbol branch of the array element;

[0050] Step 1-6: Merge the metric soft value information of the same symbol branch in all array element channels to obtain the new soft value information γ k ;

[0051] The formula is:

[0052]

[0053] Step 8: The new soft value information γ after merging k The BICM-ID method is applied for decoding to obtain the soft value information of each symbol under high-order modulation. After decoding, the information bits carried by each symbol are obtained, thereby improving the system's anti-noise performance.

[0054] Specifically:

[0055] Step 1-1: Calculate the log-likelihood ratio L(u) of each information bit according to the maximum a posteriori probability criterion. k,j ):

[0056]

[0057] Where y represents the set of all information symbols of a frame received after multi-channel merging; it is expressed as:

[0058] y={y0(k),y1(k),...,y l (k),...,y L-1 (k)};

[0059] Step 1-2: Using new soft value information γ k Solve the log-likelihood ratio to obtain the soft value information of the high-order bit modulation symbol;

[0060] Step 1-3: judge the solved soft value information and output the information bit value;

[0061] The judgment rules are:

[0062]

[0063] The advantages of the present invention are:

[0064] 1) The present invention provides a joint processing method for MPSK demodulation and Doppler tracking for underwater acoustic direct spread system, which realizes time-varying Doppler tracking, estimation and compensation in the context of high-order modulation through a survivor-by-survivor algorithm and symbol-by-symbol velocity estimation;

[0065] 2) The present invention provides a combined MPSK demodulation and Doppler tracking processing method for an underwater acoustic direct-spread system, which utilizes double differential modulation technology to resist constant Doppler interference and improve the stability of the receiving system;

[0066] 3) The present invention provides a joint MPSK demodulation and Doppler tracking processing method for underwater acoustic direct-spread system, which utilizes bit-interleaved coding and iterative decoding technology to eliminate the noise amplification effect caused by differential modulation, enabling the effective application of differential modulation technology to practical systems.

[0067] 4) The present invention provides a joint processing method for MPSK demodulation and Doppler tracking for underwater acoustic direct spread systems, which utilizes a multi-channel soft value merging method to achieve multi-channel joint processing under time-varying channels and improve receiver processing performance.

[0068] 5) The present invention provides a joint processing method for MPSK demodulation and Doppler tracking for underwater acoustic direct-spread system, which not only ensures the robust communication of direct-spread high-order modulation, but also can perform accurate Doppler tracking and estimation under time-varying channels, thereby realizing the joint communication and measurement capability under underwater acoustic channels. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 The present invention is a flow chart of the MPSK demodulation and Doppler tracking joint processing method for underwater acoustic direct spread system.

[0070] Figure 2 This is a frame structure diagram of the signal transmitted in the present invention.

[0071] Figure 3 This is a block diagram of the receiver used in the present invention.

[0072] Figure 4 This is a diagram of the decoding performance under a time-varying channel in the present invention.

[0073] Figure 5 This is a diagram of the decoding performance of multiple channels in the present invention. DETAILED DESCRIPTION

[0074] To facilitate understanding and implementation of the present invention by those skilled in the art, the present invention is further described below in detail with reference to the accompanying drawings and embodiments. It is apparent that the embodiments described are merely partial embodiments of the present invention, not all embodiments. All other embodiments derived by those skilled in the art based on the embodiments of the present invention without creative effort shall fall within the scope of protection of the present invention.

[0075] The present invention discloses a joint processing method of MPSK demodulation and Doppler tracking for an underwater acoustic direct-spread system, aiming to realize an integrated communication and measurement technology for underwater acoustics through joint communication and measurement, so that the traditional direct-spread system can achieve robust measurement capabilities while completing its communication mission.

[0076] The present invention eliminates the impact of time-varying Doppler on the communication system by combining double differential modulation technology and accurate Doppler tracking, estimation and compensation methods under high-order bit modulation interference, and adopts bit interleaving coding iterative decoding technology to reduce the impact of noise amplification brought by double differential, thereby improving the system's anti-noise performance. Finally, a novel multi-channel method is adopted to achieve robust multi-channel processing under time-varying channels by merging multi-channel soft values. This new joint processing technology not only increases the rate of traditional direct-spread communication, but also achieves high-precision measurement performance while ensuring communication robustness, and completes Doppler tracking, estimation and compensation for multiple time-varying channels. The application of this technology makes it possible to integrate underwater acoustic communication and measurement, greatly improving the efficiency of underwater acoustic operations, and has important practical application value and broad prospects.

[0077] The MPSK demodulation and Doppler tracking joint processing method for underwater acoustic direct spread system is as follows: Figure 1 As shown in the figure, the cross-correlation characteristics of the DS-SPSS signal are used to perform Doppler tracking symbol by symbol, realizing time-varying Doppler estimation and compensation within a frame of signal. Combined with the coding and decoding technology, robust communication of the DS-SPSS system under dynamic channels is achieved. The specific steps are as follows:

[0078] Step 1: Multi-channel array elements based on the ultra-short baseline array simultaneously receive the same underwater direct sequence spread spectrum communication signal and perform pre-processing;

[0079] Preprocessing includes bandpass filtering and downsampling;

[0080] like Figure 2 As shown in the figure, the frame structure of the underwater direct spread communication signal includes a synchronization header signal and several groups of data. The synchronization header is modulated by a spread spectrum code, and each group of data is modulated by other spread spectrum codes of the same length.

[0081] Step 2: For the first array element, perform synchronization capture processing on the synchronization header of the received direct spread communication signal, and use a parallel filter bank to search the starting position and coarse Doppler velocity estimate of the array element received signal in the delay Doppler two-dimensional plane.

[0082] There are L array elements, with values ​​ranging from 0 to L-1; n groups of reference signals s at different Doppler speeds are generated locally n(t), for the lth array element, the received signal is matched and correlated with the n synchronization header reference signals, the waveform with the highest peak value among the n sets of correlation results is found, and the time τ corresponding to the peak point of the waveform is output l,0 And the Doppler velocity value of the reference signal at this time

[0083] Specifically:

[0084] Step 1-1, set the Doppler speed search interval Δυ and the maximum Doppler speed υ M , and then generate a parallel filter bank containing different Doppler velocities. The Doppler velocity vector υ of the filter bank is expressed as:

[0085] υ={-υ M ,-υ M +Δυ,…,0,…,υ M -Δυ,υ M}

[0086] Step 1-2: Generate n sets of local complex passband reference signals s with different pulse widths using the Doppler velocity vector υ n (t), the received signal of the lth array element is compared with the n-group reference signal s n (t) is matched and correlated, and the absolute value is taken to obtain the envelope of each branch, and then the delay-Doppler two-dimensional correlation matrix R(τ,υ) is obtained.

[0087] Step 1-3: Calculate the decision threshold based on the mean and variance of the signal noise segment and use it to detect whether the current signal contains the expected signal; if so, proceed to step 1-4; otherwise, update the signal and return to step 1-2.

[0088] Step 1-4: In the delay-Doppler two-dimensional plane, use the matrix R(τ,υ) to search for the highest peak value among the n groups of correlation results, and then obtain the delay τ corresponding to the received signal of the array element. l,0 , according to the time delay, the starting position of the lth array element receiving the spread spectrum signal is determined, and the desired communication signal r is intercepted according to the value l (t) and obtain the initial rough Doppler velocity estimate

[0089] Step 3: Remove the carrier of each channel signal separately, and intercept the communication signal r according to the starting position of the signal received by the lth array element. l (t), perform down-conversion processing to obtain the baseband signal y l (t);

[0090] communication signalr l The expression of (t) is:

[0091]

[0092] where v k is the information bit of the kth symbol, A l,k is the signal amplitude of the kth symbol of the lth array element signal, υ l,k is the Doppler velocity value of the symbol, f c is the carrier center frequency of the signal, T s is the duration of a symbol, N d is the number of symbols in a frame signal, the operation symbol represents the real part, n(t) is the noise of the passband signal, and c1(t) is the despreading code of the local baseband information bit.

[0093] Down-conversion operation on communication signals l (t) is multiplied by the carrier and passed through a low-pass filter to obtain the baseband signal y l (t):

[0094]

[0095] Where LPF(·) represents low-pass filtering of the received signal, υ l,k is the Doppler velocity value of the kth symbol of the lth channel, c is the speed of sound in water, ω l (t) represents the noise of the baseband received signal.

[0096] Step 4: Based on the rough Doppler velocity estimate of the lth array element signal and baseband signal y l (t) despreads the information bits of the spreading code c1(t) to generate the local baseband reference signal

[0097] Where c is the speed of sound in water;

[0098] Step 5: Based on local baseband reference signal With the baseband signal y l (t) Perform matching correlation and despread the correlation peak of each symbol in the lth array element received signal to obtain the despread correlation signal R l (τ);

[0099]

[0100] Step 6: According to the correlation signal R of the lth array element l (τ), extract the highest peak amplitude and phase of each symbol and calculate the received signal corresponding to each symbol of the array element;

[0101] Using the maximum energy criterion, the highest peak of each symbol is selected as the position corresponding to the information symbol, and the information carried by the peak is selected.

[0102]

[0103] For the kth symbol of the array element, extract its highest peak amplitude and phase information Get the received symbol of the array element k∈{1,2,…,N d +2};

[0104] Step 7: The traditional PSP (Per-survivor processing) algorithm based on Viterbi algorithm is replaced with the BCJR algorithm for improvement. The Doppler shift is estimated and compensated on the branch metric corresponding to each symbol. The branch metric soft value information of the same symbol of all elements is merged to obtain the new soft value information γ k ;

[0105] The specific process is:

[0106] Step 1-1: According to the improved PSP algorithm, traverse the state quantity corresponding to each branch on the Trellis graph to obtain the symbol information bit corresponding to each state;

[0107] The received signal is processed symbol by symbol. The length of each processing unit is 2 adjacent symbols, and the update length of the processing unit is 1 symbol. The information bit corresponding to the k-th symbol state is b k , the processing unit is information bit b k and b k-1 .

[0108] Step 1-2: For the kth symbol, use the information bits and the received signal y processed symbol by symbol l (k), estimate the Doppler phase offset for this symbol time for:

[0109]

[0110] Step 1-3: Use the BCJR (Bahl-Cocke-Jelinek-Raviv) algorithm to compensate the Doppler phase offset of the k-th branch metric. Get the metric soft value information γ of the branch l,k ;

[0111] Calculated as:

[0112]

[0113] where u k =[u k,1 ,...,u k,j ,...,uk,J ],u k,j ∈{1,-1} represents the information bits transmitted; and u k satisfy:

[0114]

[0115] where j = 1,…,J; b k is a double-difference coding sequence that satisfies

[0116] N is the number of multi-symbols used, L a (u k,j ) is the symbol u k,j Prior information, M is the modulation order of multi-level phase shift, is the variance of the noise.

[0117] Step 1-4: Similarly, obtain the metric soft value information of each symbol branch of the l-th array element;

[0118] Step 1-5: Select the next array element and repeat the above steps to obtain the metric soft value information of each symbol branch of the array element;

[0119] Step 1-6: Merge the metric soft value information of the same symbol branch in all array element channels to obtain the new soft value information γ k ;

[0120] The formula is:

[0121]

[0122] Step 8: The new soft value information γ after merging k The BICM-ID (Bit-interleaved coded modulation with iterative decoding) method is applied for decoding to obtain the soft value information of each symbol under high-order modulation. After decoding, the information bits carried by each symbol are obtained, improving the system's noise resistance.

[0123] Specifically:

[0124] Step 1-1: Calculate the log-likelihood ratio L(u) of each information bit according to the maximum a posteriori probability criterion. k,j ):

[0125]

[0126] Where y represents the set of all information symbols of a frame received after multi-channel merging; it is expressed as:

[0127] y={y0(k),y1(k),...,y l (k),...,y L-1 (k)};

[0128] Step 1-2: Using new soft value information γ k Solve the log-likelihood ratio to obtain the soft value information of the high-order bit modulation symbol;

[0129] The specific process is:

[0130] First, using the Bayesian criterion, we get

[0131] Because no matter for +1 or -1, there are P(y) above and below, after canceling, P(u k,j ,y) to solve: For the Trellis graph, when the current state and input bit symbol information are given, the state at the next moment can be determined. Therefore, the state at the current k symbol moment is defined as s′, and the state at the next moment is s, and we get:

[0132] in Indicates that for u k,j = +1 state pairs s and s'; the above formula can be equivalent to:

[0133] P(s',s,y)=P(y t＞k |s)P(s,y k |s')P(s',y t＜k )

[0134] The above formula is mainly derived from the Bayesian criterion and the Markov chain simplification. The Markov chain criterion states that when a random process is given the current state and all past states, the conditional probability distribution of the future state depends only on the current state. Based on the above formula, we define:

[0135] α k (s')≡P(s',y t＜k )

[0136] γ k (s',s)≡P(s,y k |s')

[0137] β k+1 (s)≡P(y t＞k |s)

[0138] Therefore, P(s',s,y) can be written as: P(s',s,y)=α k (s')γ k (s',s)β k (s);

[0139] The forward metric and the backward metric can be calculated from the branch metric:

[0140]

[0141] Where σ k is the set of all states at time k.

[0142] Therefore, we can use forward recursion to calculate the forward metric α for each state s at time k+1 k+1 (s). Similarly, the probability β k The expression of (s′) is written as above:

[0143]

[0144] And the backward recursion can be used to calculate the backward metric β of each state s′ at time k k (s′).

[0145] The forward recursion starts at time k=0, with the initial condition

[0146]

[0147] So the encoder starts from the all-zero state. Similarly, the backward recursion starts from the last bit with the initial condition

[0148]

[0149] Recursively calculate β k (s).

[0150] Step 1-3: judge the solved soft value information and output the information bit value;

[0151] The judgment rules are:

[0152]

[0153] Furthermore, the symbol u k,j Prior information L a (u k,j ), when the first iteration takes a value of 0, when the number of iterations is greater than 1, the next prior information L a (u k,j ) is the log-likelihood ratio of the decision information of the previous symbol

[0154] Example:

[0155] The MPSK demodulation and Doppler tracking joint processing method for underwater acoustic direct spread system proposed in the present invention was simulated and verified. The simulation was performed on a five-channel receiving array. The basic signal parameters are shown in Table 1:

[0156] Table 1

[0157] parameter set up communication system direct expansion communication Modulation method 8PSK Operating frequency band 11.25-13.75kHz Center frequency 12.5kHz Spread spectrum code length of synchronization header N＝511 Spreading code length of information bits N＝127 Spreading code for synchronization header and information bits m-sequence

[0158] The following steps are involved:

[0159] Step 1: Use multiple array elements to receive underwater direct-spread communication signals;

[0160] The receiving end uses multiple array elements to collect the signals sent by the transmitting end. Each array element receives the same direct-spread signal, which has slight delay differences and different noise interference. The received signal is pre-processed to reduce noise and interference, improve signal quality, and lay the foundation for subsequent delay estimation and signal processing.

[0161] Step 2: Each array element performs signal synchronization and Doppler coarse estimation.

[0162] Step 3: Remove the carrier of each channel signal.

[0163] Step 4: Despread the passband signal.

[0164] Step 5: Extract the signal peak information of the kth symbol of the lth array element.

[0165] Step 6: Perform fine Doppler estimation and compensation for each channel.

[0166] Step 7: Merge multi-channel soft value information.

[0167] Step 8: Iterative decoding.

[0168] Steps 2 to 8 are all processing steps of the receiver, and their connections can be seen Figure 3 The receiver block diagram used is shown.

[0169] Under the above signal parameters, a time-varying underwater acoustic channel is constructed to simulate the uniformly accelerated motion of the transmitter and receiver. The system communication performance of the receiver proposed in this invention is traversed under different acceleration backgrounds and compared with the traditional method. The traditional method is the time-varying Doppler velocity estimation method used in Reference 2. The simulated SNR-BER curve is shown as follows: Figure 4 shown.

[0170] As can be seen from the figure, when the system acceleration is less than 0.61m / s 2The receiver of the present invention outperforms the traditional receiver and has a performance gain of more than 5dB. The performance of the receiver of the present invention with different numbers of array elements is simulated in a multi-channel background. The system is accelerated at 0.14m / s 2 Under this background, the simulated SNR-BER curve is as follows Figure 5 As shown, multi-channel combining has significant processing gain.

[0171] The signal parameters for the field test are a center frequency of 12.5kHz and a frequency band of 10-15kHz. The transmitted signal frame format is the same as the test, but the spread spectrum code is 511. The field test contains a total of 250 frames of 8PSK data. The signal decoding effect is shown in Table 2 below.

[0172] Table 2

[0173] Treatment method Correctly decoded frames Number of frames decoded incorrectly The receiver proposed by the present invention 249 1 Traditional receiver[2] 239 11

[0174] The results show that the present invention improves the processing performance of traditional direct-spread receivers under high-order modulation and time-varying Doppler background by finely tracking, estimating and compensating the Doppler offset on each state branch through an improved survival-by-survivor algorithm and state transition diagram, merging multi-channel soft values ​​and estimating the maximum a posteriori likelihood ratio of symbols using the BCJR algorithm. It can not only achieve robust communication, but also high-precision Doppler measurement, making it possible to integrate underwater acoustic communication and measurement, greatly improving the efficiency of underwater acoustic operations, and has significant technical advantages and broad application prospects.

[0175] The present invention uses a matched filter group to detect the received signal. When the signal is detected, a peak search is performed in a two-dimensional plane of delay Doppler frequency to achieve signal synchronization and initial Doppler estimation of the signal. A local passband reference signal is generated using the initial Doppler estimation result, and the starting position of the spread spectrum signal is determined according to the synchronization position. The received signal is intercepted and down-sampled, the passband signal is moved to the baseband and the carrier is removed, and then correlation processing is performed with the local baseband reference signal to obtain a baseband correlation signal. The amplitude and phase of the highest peak of the correlation signal corresponding to each symbol are extracted, and thus the soft information of the baseband communication symbol is obtained. Continuous Doppler tracking is then performed, and the Doppler offset estimation value on the branch metric corresponding to each state is obtained and compensated according to an improved survival-by-survivor algorithm and a state transition diagram. The branch metrics after compensation of each channel are combined into multiple channel soft values, and the likelihood ratio of the combined symbol is estimated using the BCJR algorithm. Finally, bit interleaved coding iterative decoding is used to realize double differential soft demodulation, and multiple iterations are performed to improve the anti-noise performance of the system. The present invention improves the processing performance of the high-order modulated direct-spread system in the underwater acoustic time-varying channel, and while ensuring stability, increases the rate of underwater acoustic direct-spread communication, making the traditional underwater acoustic direct-spread communication machine more applicable, with a longer operating range and higher robustness. In addition, the present invention combines communication and measurement technologies to perform precise Doppler tracking and estimation in the time-varying channel while achieving communication, thereby realizing the joint communication and measurement capability in the underwater acoustic channel.

Claims

1. A combined MPSK demodulation and Doppler tracking processing method for underwater acoustic direct spread system, characterized in that: The following steps are involved: Step 1: Multi-channel array elements based on the ultra-short baseline array simultaneously receive the same underwater direct-spread communication signal and perform pre-processing; Step 2: For the lth array element, the synchronization header of the received DSSS communication signal is synchronously captured and processed, and a parallel filter bank is used to search the starting position and coarse Doppler velocity estimate of the array element received signal in the delay-Doppler two-dimensional plane. There are L array elements in total, with values ​​ranging from 0 to L-1; Step 3: The communication signal r intercepted based on the starting position of the signal received by the lth array element l (t), perform down-conversion processing to obtain the baseband signal y l (t); Step 4: Based on the rough Doppler velocity estimate of the lth array element signal and baseband signal y l (t) despreads the information bits of the spreading code c1(t) to generate the local baseband reference signal Step 5: Based on local baseband reference signal Despread the correlation peak of each symbol in the received signal of the lth array element to obtain the despread correlation signal R l (τ); Step 6: According to the correlation signal R of the lth array element l (τ), obtain the highest peak amplitude and phase of each symbol, and calculate the received signal corresponding to each symbol of the array element; For the kth symbol of the array element, extract its highest peak amplitude and phase information Get the received symbol of the array element Step 7: The traditional PSP algorithm is improved by replacing the Viterbi algorithm with the BCJR algorithm. The Doppler shift is estimated and compensated on the branch metric corresponding to each symbol. The branch metric soft value information of the same symbol of all elements is combined to obtain the new soft value information γ k ; The specific process is: Step 1-1: According to the improved PSP algorithm, traverse the state quantity corresponding to each branch on the Trellis graph to obtain the symbol information bit corresponding to each state; The information bit corresponding to the kth symbol state is b k ; Step 1-2: For the kth symbol, use the information bits and the received signal y processed symbol by symbol l (k), estimate the Doppler phase offset for this symbol time for: Step 1-3: Compensate the Doppler phase offset of the k-th branch Get the metric soft value information γ of the branch l,k ; Calculated as: where u k =[u k,1 ,...,u k,j ,...,u k,J ],u k,j ∈{1,-1} represents the information bits transmitted; and u k satisfy: where j = 1,…,J; b k is a double-difference coding sequence that satisfies N is the number of multi-symbols used, L a (u k,j ) is the symbol u k,j The prior information of is 0 when it is first iterated, M is the modulation order of the multi-level phase shift, is the variance of the noise; Step 1-4: Similarly, obtain the metric soft value information of each symbol branch of the l-th array element; Step 1-5: Select the next array element and repeat the above steps to obtain the metric soft value information of each symbol branch of the array element; Step 1-6: Merge the metric soft value information of the same symbol branch in all array element channels to obtain the new soft value information γ k ; The formula is: Step 8: The new soft value information γ after merging k The BICM-ID method is applied for decoding to obtain the soft value information of each symbol under high-order modulation. After decoding, the information bits carried by each symbol are obtained, thereby improving the system's anti-noise performance.

2. The MPSK demodulation and Doppler tracking joint processing method for underwater acoustic direct spread system according to claim 1, characterized in that: In the step 1, the preprocessing includes bandpass filtering and downsampling; The frame structure of the underwater direct spread communication signal includes a synchronization header signal and several groups of data. The synchronization header is composed of spread spectrum code modulation, and each group of data is composed of other spread spectrum code modulations of the same length.

3. The MPSK demodulation and Doppler tracking joint processing method for underwater acoustic direct spread system according to claim 1, characterized in that: The step 2 is specifically as follows: Step 1-1, set the Doppler speed search interval Δυ and the maximum Doppler speed υ M , and then generate a parallel filter bank containing different Doppler velocities. The Doppler velocity vector υ of the filter bank is expressed as: υ={-υ M ,-y M +Dy,…,0,…,y M -Yes, yes M } Step 1-2: Generate n sets of local complex passband reference signals s with different pulse widths using the Doppler velocity vector υ n (t), the received signal of the lth array element is compared with the n-group reference signal s n (t) Perform matching correlation and take the absolute value to obtain the envelope of each branch, and then obtain the delay-Doppler two-dimensional correlation matrix R(τ,υ); Step 1-3: Calculate the decision threshold based on the mean and variance of the signal noise segment and use it to detect whether the current signal contains the expected signal; if so, proceed to step 1-4; otherwise, update the signal and return to step 1-2; Step 1-4: In the delay-Doppler two-dimensional plane, use the matrix R(τ,υ) to search for the highest peak value among the n groups of correlation results, and then obtain the delay τ corresponding to the received signal of the array element. l,0 , according to the time delay, determine the starting position of the lth array element receiving the spread spectrum signal, and intercept the desired communication signal r according to this value l (t) and obtain the initial rough Doppler velocity estimate 4. The MPSK demodulation and Doppler tracking joint processing method for underwater acoustic direct spread system according to claim 1, characterized in that: In the step 3, the communication signal r l The expression of (t) is: where v k is the information bit of the kth symbol, A l,k is the signal amplitude of the kth symbol of the lth array element signal, υ l,k is the Doppler velocity value of the symbol, f c is the carrier center frequency of the signal, T s is the duration of a symbol, N d is the number of symbols in a frame signal, the operation symbol represents the real part, n(t) is the noise of the passband signal; c1(t) is the despreading code of the local baseband information bit; Down-conversion operation on communication signals l (t) is multiplied by the carrier and passed through a low-pass filter to obtain the baseband signal y l (t): Where LPF(·) represents low-pass filtering of the signal, c is the speed of sound in water, ω l (t) represents the noise of the baseband received signal.

5. The MPSK demodulation and Doppler tracking joint processing method for underwater acoustic direct spread system according to claim 1, characterized in that: The step eight is specifically as follows: Step 1-1: Calculate the log-likelihood ratio L(u) of each information bit according to the maximum a posteriori probability criterion. k,j ): Where y represents the set of all information symbols of a frame received after multi-channel merging; it is expressed as: y={y0(k),y1(k),...,y l (k),...,y L-1 (k)}; Step 1-2: Using new soft value information γ k Solve the log-likelihood ratio to obtain the soft value information of the high-order bit modulation symbol; Step 1-3: judge the solved soft value information and output the information bit value; The judgment rules are:

Images