A multi-peak doppler estimation compensation method based on an OCDM underwater acoustic communication system

Abstract

The application discloses a multi-peak Doppler estimation compensation method based on an OCDM underwater acoustic communication system and relates to the field of underwater acoustic communication. A string of known sequences is inserted before and after each OCDM signal block for Doppler estimation and compensation; through the OCDM signal of the SSML channel model, firstly, oversampling operation is carried out, then, autocorrelation operation is used to generate multiple peak values, after judging whether the Doppler scale has a relatively large change during signal transmission, segmented processing is adopted for the relatively large change interval; in each segment, the multiple peak values are used to estimate and compensate the Doppler scaling factor, after performing downsampling operation, residual Doppler is further processed, so that high-precision Doppler estimation effect is obtained, and the accuracy of the OCDM signal in underwater acoustic communication is improved. The application fully utilizes the design structure of the OCDM system transmission frame in the estimation of the Doppler frequency shift of the underwater acoustic channel, improves the Doppler estimation precision while ensuring the bandwidth utilization, and enhances the anti-interference ability and communication reliability of the system.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of underwater acoustic communication, and particularly relates to a multi-peak Doppler estimation compensation method based on an OCDM underwater acoustic communication system. BACKGROUND

[0002] With the emergence of the Orthogonal Chirp Division Multiplex (OCDM) modulation technology in recent years, the technology uses a series of mutually orthogonal chirp signals to carry information, and the digital representation of the technology is a Fresnel transform, so as to realize the mutual change of the signal time domain and the Fresnel domain. The information transmission of OCDM shows good anti-interference in the time-frequency domain, and is gradually used in optical communication, satellite communication and other scenes. However, due to the characteristics of the underwater communication environment with large Doppler scale, it is necessary to estimate the Doppler scaling factor of the received signal. At present, the method for estimating the Doppler factor of the OCDM system underwater acoustic communication is to add a preamble and a postamble before and after the transmitted signal, but the accuracy of such estimation method is difficult to improve. In addition, the existing algorithm uses the method of transmitting pilots first and then transmitting multiple OCDM information blocks, and channel estimation cannot be realized within the information block. When the channel time-varying is serious, if effective communication is to be realized, the length of the transmission frame will be limited. Moreover, many Doppler estimation methods for Orthogonal Frequency Division Multiplex (OFDM) underwater acoustic communication systems cannot be directly applied to OCDM systems, highlighting the necessity of designing an adaptive Doppler estimation method for OCDM systems. In addition, if the transmission signal cannot maintain relative static state between the transmitting and receiving ends, and the two relatively move at a large speed, the correct Doppler scaling factor cannot be estimated, resulting in failure of signal demodulation. Therefore, it is necessary to consider the factors that cause the Doppler effect under water and design a Doppler estimation method based on the OCDM underwater acoustic communication system.

[0003] The application number 201710608131.1 entitled "OFDM-MFSK underwater acoustic communication wideband Doppler estimation and compensation method of known subcarrier frequency" provides an underwater acoustic communication wideband Doppler estimation method for OFDM system, which can complete accurate estimation and compensation of Doppler in OFDM system through one-time oversampling. The application number 201711148334.3 entitled "Doppler factor estimation and compensation method for mobile underwater acoustic communication" provides a low-complexity Doppler estimation method for fast-changing underwater acoustic communication signals. The application number 202011171021.1 entitled "Doppler estimation and compensation method for multi-carrier underwater acoustic communication based on motion platform" provides a Doppler estimation and compensation method for motion platform multi-carrier underwater acoustic communication with time-frequency domain insertion of transmission factors. The above methods solve the Doppler compensation and estimation of OFDM system in underwater acoustic communication, but due to the difference in subcarrier form of the two systems, they cannot be directly applied to OCDM underwater acoustic communication system. At present, the Doppler estimation method used in OCDM system is still the concept of front and rear code estimation, which limits the estimation accuracy and anti-interference ability of OCDM system in underwater acoustic communication.

[0004] Therefore, the skilled in the art is committed to developing a multi-peak Doppler estimation and compensation method based on OCDM underwater acoustic communication system. The present application faces OCDM underwater acoustic communication system and proposes a multi-peak Doppler estimation and compensation method for OCDM signals. Different from the traditional Doppler scaling factor estimation method, the present application uses Single Scale Multi Lag (SSML) to simulate the underwater acoustic communication environment, comprehensively considers the influence of Doppler effect in underwater acoustic communication and whether the relative speed of the transmitting and receiving ends changes significantly during transmission, and designs a high-precision Doppler estimation and compensation method for OCDM underwater acoustic communication system. By using the characteristics of each block in the designed OCDM transmission frame that can generate multiple autocorrelation peaks, it is determined whether the relative speed of the transmitting and receiving ends changes rapidly during signal transmission, and then the Doppler scaling factor is estimated and compensated, and then the residual Doppler is processed. This method reduces the bandwidth utilization while achieving high-precision estimation and compensation of OCDM received signals. SUMMARY

[0005] In view of the above defects of the prior art, the technical problem to be solved by the present application is to solve the signal restoration problem of OCDM underwater acoustic communication system from the underwater acoustic channel model with Doppler effect, and to solve the current situation that the Doppler estimation accuracy is limited and the system anti-interference ability is poor.

[0006] To achieve the above purpose, the present application provides a multi-peak Doppler estimation and compensation method based on OCDM underwater acoustic communication system, comprising the following steps:

[0007] Step 1, generating the sending frame in the sending terminal of the OCDM underwater acoustic communication system, containing pilot symbols v1(i), v2(i) with length L, data symbols d(i) with length N Wherein is the inverse Fresnel transform matrix, d(i) is the data information to be transmitted, and the combination is an OCDM signal block with length M = 2L + M K OCDM signal blocks are generated in total, after a serial-parallel conversion operation, a pre- and post- code v pre ,v post is added respectively to allocate energy and bandwidth to the above components;

[0008] Step 2, using a pulse shaper p(t) on the sending frame to obtain a continuous passband signal

[0009] Step 3, transmitting the passband signal through a power amplifier, a matching circuit and a transducer, converting the electrical signal into an acoustic signal and emitting it to the underwater environment;

[0010] Step 4, using the SSML channel model h(t,τ) to simulate the underwater communication environment, using a to represent the Doppler scaling factor, A p and τ p are the channel gain and time delay respectively, is the passband additive white Gaussian noise;

[0011] Step 5, the receiving signal of the OCDM underwater acoustic communication system sending signal after passing through the SSML channel

[0012] Step 6, performing an oversampling operation on the received passband signal at an oversampling rate λ to obtain a discrete passband receiving signal performing an autocorrelation operation, with a template being a sequence of length λL after the synchronization point of the receiving signal, is the search range related to the Doppler scaling factor in each block, thereby obtaining multiple Doppler-related peaks and then using a threshold discrimination method Wherein Γ th is the threshold used for discrimination, and the Doppler scaling factor calculated for each single peak is judging whether the relative speed between the sending terminal and the receiving terminal during the sending signal has changed greatly, if so, segment processing is needed, and the starting block index K1 and the ending block index K2 are determined;

[0013] Step 7, Estimate the Doppler scaling factor using multi-peaked Doppler estimation algorithm in the interval i = K1, …, K2 To the discrete passband received signal Perform resampling operation, preliminary compensate the Doppler scaling factor, obtain the resampled OCDM received signal Then perform down-sampling de-carrier operation, obtain the OCDM baseband discrete signal y[n], which contains residual Doppler w o .

[0014] Step 8, Perform serial-parallel conversion operation to the resampled signal, obtain the i-th signal block expression y(i) using the spliced new sequence Estimate the CFO parameter in the i-th block, where is the symbol length without interference, write the closed expression of CFO in the signal And compensate the signal block, obtain the compensated signal z(i);

[0015] Step 9, Perform channel estimation operation to the signal z(i) removing the Doppler effect, obtain the channel estimator And Perform pre-processing operation, including removing the tail of adjacent sequence and superimposing the tail of itself, obtain the pre-processed signal Perform equalization operation, obtain the OCDM symbol of the received signal estimation

[0016] Further, the step 1, the OCDM signal block is modulated by a Discrete Fresnel Transform (DFnT) matrix, the (m, n) input of the DFnT matrix with size N × N is Where N is the number of subcarriers of the OCDM system. The OCDM structure frame is composed of payload information And two known sequences v1(i) and v2(i) with length L. The data symbol vector with length N is taken from the complex modulation alphabet, denoted as d(i) = [d(iN), d(iN+1), …, d(iN+N-1)] T . The subsequent derivation is completed with even N, and odd N has similar derivation process, which is not described here. Where v1(i) and v2(i) are defined as The pilot symbol is designed as p(i) = [a, 0, …, 0] T , and the fixed power is respectively allocated to the two known sequences. The i-th transmitted signal block in the transmitted frame has a length of M = 2L + N, which can be written as The pilot needs to cover the whole bandwidth B, and the frequency resolution of the three parts of the transmission block s(i) is and After the serial-to-parallel conversion, one known sequence v of length L is added at the beginning and at the end pre ,v post The two known sequences are also defined as The nth input to generate the baseband transmission signal is denoted as s[n], which is specifically The energy E is used to transmit the OCDM structure frame, which contains K transmission blocks. Let the pilot and the symbols in the transmission data experience the same signal-to-noise ratio, and the energy allocated to the transmission data is The two parts of the pilot in the structure frame can each be allocated to Therefore, the energy allocated to the pilot is

[0017] Further, the step 6 comprises the following steps:

[0018] Step 6.1, performing an oversampling operation on the passband OCDM received signal, and the oversampling rate is where T s is the sampling frequency of the transmitted signal, T r is the sampling frequency of the received signal, and a discrete passband received signal

[0019] Step 6.2, performing an autocorrelation operation on the discrete passband received signal to synchronize the first chirp signal in the first block after the first chirp signal is used as a template, and a plurality of Doppler correlation peaks where the Doppler search subset is related to the maximum scale change and contains [-α max (λL+λMi), α max (λL+λMi)], where α max is the maximum Doppler scale;

[0020] Step 6.3, using a threshold discrimination method where the Doppler scaling factor calculated by the single peak is Γ th is a threshold value for determining whether the received signal needs to be segmented. The starting block index K1 and the ending block index K2 of the subsequent operation are determined.

[0021] Further, the step 7 comprises the following steps:

[0022] Step 7.1, perform the following operation between index blocks K1 and K2, for the OCDM discrete passband received signal, utilize multiple Doppler related peaks Estimate the Doppler scaling factor where K min represents the minimum interval between the starting point K1 and the K min th related peak.

[0023] Step 7.2, perform resampling operation on the OCDM passband received signal with the estimated Doppler scaling factor to obtain the resampled OCDM received signal

[0024] Step 7.3, perform down-sampling de-carrier operation on the resampled signal with a down-sampling ratio of where f c is the carrier frequency, T s is the transmitted signal sampling frequency, T r is the received signal sampling frequency, to obtain the OCDM baseband received signal y[n];

[0025] Step 7.4, assume that 1 + a and are very close, represent the OCDM baseband signal after resampling operation on the OCDM baseband received signal where residual Doppler w o ∈(0, 1), w[n] is the baseband AWGN.

[0026] Further, the step 8 comprises the following steps:

[0027] Step 8.1, perform serial-parallel conversion operation on the resampled signal, express the OCDM baseband signal y[n] in the form of matrix, where obtain the expression of the ith signal block where D N+L (w o ) represents the residual Doppler interference within the block, H is the channel parameter matrix, is the inverse Fresnel transform matrix.

[0028] Step 8.2, estimate the CFO parameter in the ith block using the spliced new sequence to write the closed expression of the CFO interference within the signal

[0029] ​Step 8.3, using the estimated CFO value to resample the received signal Compensate the CFO in the signal, if the CFO in the signal is completely eliminated, the sub-carrier in the compensated signal is orthogonal, and the compensated signal z(i) is obtained.

[0030] Further, the step 9 includes the following steps:

[0031] Step 9.1, performing channel estimation operation on the signal after removing the Doppler effect, and obtaining the channel estimator When the channel variance cannot be obtained, use

[0032] Step 9.2, performing pre-processing operation on the signal after removing the Doppler effect, and first removing the tail interference of the previous adjacent chirp signal from the OCDM signal block in the pre-processing process, which is represented as Then, the tail of the OCDM signal block is added to the received OCDM signal block by the cyclic superposition operation, and then the left multiplication matrix F N Shift the signal to the frequency domain, and the expression of the pre-processed signal is

[0033] Step 9.3, performing equalization operation on the pre-processed signal Performing equalization operation on the pre-processed signal Where Γ N is the diagonalized Fresnel transform parameter matrix, and the ZF equalizer is G ZF (i) = Λ -1 (i), where the diagonal matrix is the channel frequency response, and R OLA is the cyclic superposition matrix, and the MMSE equalizer is Where is the power of the transmitted data symbol in the structure frame, and the additive noise variance matrix is Where σ 2 is the noise variance.

[0034] Further, the underwater acoustic communication machine transmitting end includes a digital-to-analog conversion module, a power amplifier, a matching circuit, and a transducer.

[0035] Further, the underwater acoustic communication machine receiving end includes an acoustoelectric conversion module, a signal amplification module, and a band-pass filter.

[0036] Further, the steps 1 to 3 are the underwater acoustic communication machine transmitting end operations.

[0037] Further, the steps 5 to 9 are the underwater acoustic communication machine receiving end operations.

[0038] In the preferable embodiment of the present application, the existing OCDM system uses the traditional Doppler estimation method of adding pre / post-codes, which cannot further improve the Doppler estimation accuracy, and the underwater acoustic signal communication effect is limited, and the anti-interference performance is poor. For the OCDM underwater acoustic communication system, a more reasonable sending frame structure is designed, so that the received signal can obtain multiple correlation peaks, and the Doppler scaling factor is estimated and the residual Doppler is processed, which improves the estimation accuracy of the Doppler and the anti-interference ability, and further improves the underwater communication effect of the OCDM signal. Through the design of the sending frame of the OCDM system, multiple correlation peaks can be generated after the autocorrelation of the received signal, after using the threshold judgment method to judge whether it needs to be segmented, the multi-peak Doppler estimation compensation method is used to process the signal, and then the closed expression of the residual Doppler estimation is derived by using the subcarrier pilot characteristics in the OCDM signal frame, and then the compensation is performed.

[0039] The existing OCDM underwater acoustic communication system does not consider the OCDM signal through the underwater acoustic channel model with Doppler characteristics, and does not study the influence of Doppler on the OCDM signal from the principle. The present application adopts the SSML channel model to simulate the underwater environment, and the principle derivation of the Doppler scaling factor estimation and compensation is carried out for the received signal of the OCDM underwater acoustic communication system through the SSML channel model. The two-step elimination method is used to eliminate the Doppler scaling factor of the received OCDM signal through the SSML channel, the resampling method is used for pre-processing of the received OCDM signal through reasonable approximation, and then the residual Doppler is further estimated and compensated.

[0040] The existing Doppler estimation technology is difficult to cope with the case that the relative speed of the transmitting and receiving ends suddenly changes during the sending signal, and the anti-interference ability is weak. The present application considers the time-varying channel, realizes the ability of estimating the Doppler scaling factor, residual Doppler and channel information in each OCDM block, considers the scene that the relative speed of the transmitting and receiving ends suddenly changes during the sending signal, designs a threshold discrimination algorithm, and if the detection is greater than the threshold, the received signal is segmented. The pilot signal in the OCDM block is used to complete the estimation of the Doppler scaling factor, residual Doppler and channel information, and compensation is performed, so that it can have stronger anti-interference ability.

[0041] The present application faces the present stage OCDM system in the underwater acoustic communication scene, and provides a multi-peak Doppler estimation compensation method based on the OCDM underwater acoustic communication system, aiming at the present situation that the underwater Doppler frequency shift is large, the estimation accuracy is limited, and the anti-interference ability is poor.

[0042] The application provides a Doppler estimation compensation method for an OCDM underwater acoustic communication system. The application adopts a linear time-varying channel model with a common Doppler spread, i.e., an SSML model, as a channel model to simulate an underwater acoustic communication environment, each path of the model has the same Doppler scaling factor, so that the received waveform can be uniformly compressed or stretched, and the SSML model mainly simulates the main Doppler shift of the received signal, which is caused by the direct movement between the sending end and the receiving end.

[0043] The application provides a multi-peak Doppler estimation method for an OCDM signal in an underwater acoustic communication environment, for an OCDM signal through an SSML model, first, a downsampling operation is performed, then a plurality of peak values are generated by using an autocorrelation operation, after judging whether the Doppler scale has a large change, the interval with a large change is segmented. In each segment, the plurality of peak values are used to estimate the Doppler scaling factor, then a downsampling operation is performed to further process the residual Doppler, so as to obtain a high-precision Doppler estimation effect and improve the accuracy of the OCDM signal in underwater acoustic communication. The application fully utilizes the design structure of the OCDM system sending frame to improve the Doppler estimation precision while ensuring the bandwidth utilization.

[0044] The application considers processing the received signal of the OCDM system under the influence of large Doppler shift, mainly describes the underwater acoustic communication application scenario, and the high-speed moving object on water also has the characteristics of large Doppler shift and multi-path, which can be processed in the same way.

[0045] The specific design steps of the sending end and the receiving end of the OCDM underwater acoustic communication system multi-peak Doppler estimation algorithm are as follows, the following steps 1 to 3 are sending end operations, steps 5 to 9 are receiving end operations, and in specific implementation, the sending information can be fixed after the sending end is executed once, and the steps 5 to 9 can be repeatedly executed by repeatedly sending at different times and different positions.

[0046] (1) Step 1: Generating the transmitted frame s[n] in the transmitting end of the OCDM underwater acoustic communication system, including pilot symbols v1(i), v2(i), data symbols u(i) and pre / post-codes v pre post , respectively, and assigning them energy E and bandwidth B.

[0047] The OCDM signal block is modulated by a Discrete Fresnel Transform (DFnT) matrix, and the (m, n) input of the N x N DFnT matrix is

[0048]

[0049] where N is the number of subcarriers of the OCDM system.

[0050] The structure frame is composed of load information and two known sequences v1(i) and v2(i) with length L. The data symbol vector with length N is taken from a complex modulation alphabet, denoted as d(i) = [d(iN), d(iN+1), …, d(iN+N-1)] T . The subsequent derivation is completed with even N, and odd N has a similar derivation process, which is not described here. Where v1(i) and v2(i) are defined as The pilot symbol is designed as p(i) = [a, 0, …, 0] T , and the fixed power is respectively allocated to the two known sequences. The length of the i-th transmitted signal block in the transmitted frame is M = 2L+N, which can be written as The pilot needs to cover the entire bandwidth B, and the frequency resolutions of the three parts of the transmission block s(i) are and After serial-parallel conversion, a known sequence v pre v post with length L is added at the beginning and end, and the two known sequences are also defined as The n-th input of the baseband transmission signal is denoted as s[n], which is specifically

[0051] The OCDM structure frame is transmitted with energy E, including K transmission blocks. Let the pilot and the symbols in the transmission data all experience the same signal-to-noise ratio, and the energy is allocated to the transmission data, and there are two parts of the pilot in the structure frame, which can be allocated to energy, so

[0052] (2) Step 2: Use the pulse shaper p(t) on the transmitted frame to obtain the continuous passband signal ​

[0053] The sampling rate of the pulse shaper is T s , which converts s[n] from a discrete signal to a baseband continuous signal

[0054]

[0055] where p(t) is the pulse shaper, and the bandwidth of the baseband signal is After the loading wave f c is operated, the passband signal corresponding to the structure frame is expressed as

[0056] (3) Step 3: The passband transmission signal is converted into an acoustic signal by a power amplifier, a matching circuit, and a transducer, and is transmitted to the underwater environment.

[0057] (4) Step 4: The underwater communication environment is simulated by the SSML channel model h(t,τ).

[0058] The underwater communication environment simulated by the SSML channel model includes three interference factors: Doppler scaling factor, channel gain, and channel delay. The SSML channel model is expressed as

[0059]

[0060] where τ p and A p are the transmission delay and channel gain of the pth path, P is the number of paths, and α is the Doppler scaling factor, which controls the compression and expansion scale of the signal. In addition, w(t) is an additive white Gaussian noise (AWGN) that obeys a zero-mean distribution with a variance of σ 2 .

[0061] (5) Step 5: The received signal of the OCDM underwater acoustic communication system after the transmission signal is sent through the SSML channel is expressed as

[0062] The signal received by the hydrophone through the underwater acoustic channel is converted into an electrical signal, which is enhanced after passing through a pre-amplification module and a band-pass filter to obtain the received signal of the passband OCDM underwater acoustic communication system at the receiving end of the SSML channel. The specific expression is

[0063]

[0064] where is the passband AWGN, and the symbol represents linear convolution.

[0065] (6) Step 6: Perform an oversampling operation on the received bandpass signal at an oversampling rate λ to obtain a discrete bandpass received signal Perform an autocorrelation operation to obtain a plurality of Doppler correlation peaks Determine whether the relative velocity between the transmitting end and the receiving end has changed greatly during the transmission of the signal by using a threshold discrimination method. If so, the signal needs to be processed in segments, and the starting block index K1 and the ending block index K2 are determined.

[0066] (61) Step 61: Perform an oversampling operation on the bandpass OCDM received signal to obtain a discrete bandpass received signal

[0067] Oversample the received signal at a sampling rate of T r , and the oversampling rate is After synchronization processing, the discrete bandpass received signal is represented as

[0068]

[0069] wherein is the sampling result of .

[0070] (62) Step 62: Perform an autocorrelation operation on the discrete bandpass received signal to obtain a plurality of Doppler correlation peaks

[0071] Use a matched filter on the oversampled signal to take the first chirp signal in the first block after synchronization as a template, perform an autocorrelation operation to generate a plurality of correlation peaks. The arrival time of the i-th autocorrelation peak in the bandpass received signal is

[0072]

[0073] wherein the Doppler search subset is related to the maximum scale change and contains [-α max (λL+λMi), α max (λL+λMi)], where α max is the maximum Doppler scale. It is emphasized that λL> α max N A is true, where N A = λKM is the duration from the synchronization point to the last peak.

[0074] (63) Step 63: Determine the starting block index K1 and the ending block index K2 for subsequent operations by using a threshold discrimination method.

[0075] The starting block index and the ending block index are determined. The Doppler scaling factor is calculated using adjacent correlation peaks. If it exceeds a set threshold, it means that the change is too large, i.e. the relative speed between the transmitting end and the receiving end changes rapidly during the transmission of the signal, and the signal needs to be processed in segments. The threshold determination method is shown below

[0076]

[0077] wherein Γ th is the threshold for determining whether the received signal needs to be processed in segments. If the difference between the Doppler scaling factors of two adjacent blocks is greater than the threshold, the received signal is processed in segments, and according to the block index i, the starting block index K1 and the ending block index K2 in each segment are set. If all the calculated single Doppler scaling factors are not greater than the threshold, i.e. K1=1 and K2=K.

[0078] (7) Step 7: In the interval i=K1,…,K2, the Doppler scaling factor is estimated using the multi-peak Doppler estimation algorithm The discrete passband received signal is resampled to preliminarily compensate the Doppler scaling factor, and the resampled OCDM received signal is obtained. Then, the down-sampling and carrier removal operation is performed to obtain the OCDM baseband discrete signal y[n], which contains the influence of the residual Doppler w o .

[0079] (71) Step 71: The Doppler scaling factor is estimated using multiple Doppler correlation peaks

[0080] The following operation is performed between the index blocks K1 and K2. For the OCDM discrete passband received signal, the multi-peak Doppler estimation algorithm is used to reasonably utilize the multiple peaks to estimate the Doppler scaling factor, which is specifically expressed as

[0081]

[0082] wherein K1≤K min ≤K2 is the minimum block index, representing the minimum interval between the synchronization point and the K min th correlation peak.

[0083] (72) Step 72: The resampling operation is performed on the OCDM passband received signal using the estimated Doppler scaling factor to obtain the resampled OCDM received signal

[0084] The Doppler scaling factor estimated by the multi-peak Doppler estimation algorithm ​The resampling operation is performed on the OCDM passband received signal, and the resampled OCDM received signal is denoted as

[0085]

[0086] (73) Step 73: The down-sampling de-carrier operation is performed on the resampled signal to obtain the OCDM baseband received signal y[n].

[0087] The down-sampling operation is performed on the resampled signal at a down-sampling ratio of , where T s is the sampling frequency of the transmitted signal, T r is the sampling frequency of the received signal, and the carrier f c is removed. The baseband signal can be denoted as

[0088]

[0089] where w[n] is the baseband AWGN.

[0090] (74) Step 74: Assuming that the condition is satisfied, the OCDM baseband signal y[n] after the resampling operation of the OCDM baseband received signal is denoted as o , where w

[0091] Assuming that 1+α and are very close, that is, the estimated Doppler scaling factor used in the resampling operation has high accuracy, we can denote the OCDM baseband signal as

[0092]

[0093] where is the residual Doppler, that is, the carrier frequency offset (CFO), w o ∈(0, 1).

[0094] (8) Step 8: The serial-to-parallel conversion operation is performed on the resampled signal to obtain the i-th signal block expression y(i). The new sequence obtained by splicing is used to estimate the CFO parameter in the i-th block, and the closed expression of the CFO in the signal is written as and the signal block is compensated to obtain the compensated signal z(i).

[0095] (81) Step 81: The serial-to-parallel conversion operation is performed on the resampled signal to obtain the i-th signal block expression y(i).​

[0096] After performing a serial-to-parallel conversion operation on the resampled signal, y[n] is expressed as a matrix, where The i-th signal block can be expressed as

[0097]

[0098] Where H is a Toeplitz matrix with a scale of (N+L)×N, and the first column is represented as h(l)=h(0),…,h(L) h ),0,…,0] T ,in To eliminate inter-block interference, we set L h +1≤L. Furthermore... It is a (N+L)×(N+L) diagonal matrix.

[0099] (82) Step 82: Use the newly spliced ​​sequence Estimate the CFO parameters in the i-th block and write the closed-form expression for the CFO interference within the signal.

[0100] The CFO parameter is determined by the first chirp signal [y(i)] in the i-th block. 1:L and the next in the preceding receive block a symbol To make an estimate, among which L h The maximum channel delay is given. Two sequences are concatenated to obtain a new sequence for estimating the CFO within the i-th block. The new sequence is represented as... The product of these symbol pairs is shown below

[0101]

[0102] in and Therefore, the closed-form expression for the CFO of the i-th block is as follows:

[0103]

[0104] in ln(·) and ln(·) represent the natural logarithm rule and the imaginary part, respectively. This expression has low dependence on channel information and is well adapted to underwater communication environments.

[0105] (83) Step 83: For the resampled signal, use the estimated CFO value. The CFO within the compensation signal is used to obtain the compensated signal z(i).

[0106] For the resampled signal, the CFO value calculated by the closed-form CFO is used to compensate the CFO value in the signal If the CFO is perfectly compensated, i.e. The subcarriers in each OCDM block will keep the orthogonality, and the compensated signal is denoted as

[0107]

[0108] (9) Step 9: Perform channel estimation operation on the Doppler-removed signal z(i) to obtain the channel estimator and Perform pre-processing operation to obtain the pre-processed signal Perform equalization operation to obtain the OCDM symbol of the received signal estimation

[0109] (91) Step 91: Perform channel estimation operation on the Doppler-removed signal to obtain the channel estimator and

[0110] The first chirp signal in the ith block is used to estimate the channel parameter, and the Minimum Mean-Square Error (MMSE) channel estimator of the ith block is shown as follows

[0111]

[0112] Wherein the channel variance matrix is denoted as When the channel variance cannot be obtained, the above matrix is degenerated into a Zero Force (ZF) channel estimator, denoted as

[0113] (92) Step 92: Perform pre-processing operation on the Doppler-removed signal to obtain the pre-processed signal

[0114] In the pre-processing process, the tailing interference of the preceding adjacent chirp signal is first removed from the OCDM signal block, denoted as Wherein is an LxL circulant matrix, the first column is h(0), …, h(L h ), 0, …, 0] T ; then, the tailing of the OCDM signal block is added to the received OCDM signal block through the cyclic superposition operation, and then the left multiplication matrix F N is used to transfer the signal to the frequency domain, and the expression of the pre-processed signal is

[0115] ​

[0116] wherein is a unitary matrix, R OLA =[I N [I N ] :,1:L ] in R N ] :,1:L denotes the first L columns of I N , and the size of the matrix is N x (N+L). R OLA H is a circulant matrix, which can be diagonalized by a discrete Fourier transform matrix, so the diagonal matrix is the channel frequency response. The eigenvalues of the inverse DFnT roots are is a N x N parameter matrix, and the elements on the main diagonal are

[0117] (93) Step 93: performing equalization operation on the preprocessed signal to obtain the estimated OCDM data symbol

[0118] performing equalization operation on the preprocessed signal, and the received signal estimated OCDM data symbol is represented as

[0119]

[0120] wherein G(i) is an equalization matrix, and the (l,l) diagonal element is [G(i)] l,l , and specifically, the ZF equalizer is G ZF (i) = Λ -1 (i), and the MMSE equalizer is wherein is the power of the transmitted data symbol in the structure frame, and the additive noise variance matrix is

[0121] Compared with the prior art, the present application has the following obvious essential features and significant advantages:

[0122] 1. The present application reasonably utilizes known information while ensuring the algorithm bandwidth utilization rate of the OCDM underwater acoustic communication system, designs a multi-peak Doppler estimation algorithm, improves the estimation accuracy of the Doppler in the OCDM received signal, and considers the actual situation of the underwater transceiver for adaptive compensation, and then writes a closed-form expression for the residual Doppler causing phase rotation, which reduces the dependence on underwater channel information and improves the adaptability of the system to the underwater environment. The multi-peak Doppler estimation algorithm improves the success rate and communication quality of the OCDM signal underwater communication.

[0123] 2. The OCDM underwater acoustic communication receiver demodulation framework provided by the application uses the SSML channel model to represent the OCDM received signal for the first time, and performs signal processing on this basis, considers the influence of Doppler on the OCDM underwater acoustic communication signal in principle, and reduces the influence of Doppler from two dimensions, thereby significantly improving the ability of the OCDM underwater acoustic communication system to cope with Doppler interference.

[0124] 3. The application improves the ability of the OCDM received signal to cope with the situation that the relative speed of the transmitting and receiving ends suddenly changes during signal transmission, and at the same time, under the condition of limited bandwidth, a pilot is added to each OCDM block to improve its ability to adapt to time-varying channels, thereby comprehensively improving the anti-interference ability of the OCDM underwater acoustic communication system.

[0125] The concept, specific structure and technical effects of the application will be further described below with reference to the accompanying drawings, so as to fully understand the purpose, features and effects of the application. BRIEF DESCRIPTION OF DRAWINGS

[0126] Figure 1 is the execution flowchart of a preferred embodiment of the application. DETAILED DESCRIPTION

[0127] The application will be described in more detail below with reference to the accompanying drawings, so as to make the technical content of the application clearer and easier to understand. The application can be embodied in many different forms of embodiments, and the protection scope of the application is not limited to the embodiments mentioned in the text.

[0128] The application considers processing the OCDM system received signal under the influence of large Doppler frequency shift, mainly describes the underwater acoustic communication application scenario, and the high-speed moving object on water also has the characteristics of large Doppler frequency shift and multi-path, which can be processed in the same way; since the OCDM underwater acoustic communication system adds a known sequence as a pilot in the time domain, such a pilot-added transmission frame can be processed in the same way under the influence of the multi-path channel with large Doppler frequency shift characteristics.

[0129] Figure 1 The application provides a specific implementation process of a multi-peak Doppler estimation and compensation method based on an OCDM underwater acoustic communication system. The application is aimed at the Doppler estimation problem in the current OCDM underwater acoustic communication system operating scenario, so in actual use, the specific parameters should be set according to the underwater acoustic communication machine and the transmission sea area.

[0130] The application provides a Doppler estimation compensation method for an OCDM underwater acoustic communication system.

[0131] The application provides a multi-peak Doppler estimation algorithm for an OCDM signal in an underwater acoustic communication environment, and the OCDM signal is based on an SSML model.

[0132] As shown in Figure 1 , the specific execution steps of the application are as follows:

[0133] Step 1: An OCDM signal block is modulated by a DFnT matrix, and the (m, n) input of the N*N DFnT matrix is where N is the number of subcarriers of the OCDM system. The structure frame is composed of load information and two known sequences v1(i) and v2(i) with a length of L. The data symbol vector with a length of N is taken from a complex modulation alphabet, and is represented as d(i) = d(iN), d(iN+1), …, d(iN+N-1)] T . The subsequent derivation is completed by using an even number N, and the derivation process of an odd number N is similar and is not described herein. The v1(i) and v2(i) are defined as The pilot symbol is designed as p(i) = [a, 0, …, 0] T, fixed power The two known sequences are allocated respectively. The length of the ith transmitted signal block in the transmitted frame is M = 2L + N, which can be written as The pilot needs to cover the entire bandwidth B, and the frequency resolutions of the three parts constituting the transmission block s(i) are and After the serial-parallel conversion, one known sequence v of length L is added at the beginning and the end respectively pre ,v post The two known sequences are also defined as The nth input for generating the baseband transmission signal is denoted as s[n], which is specifically The frame of the energy E transmission OCDM structure contains K transmission blocks. Let the pilot and the symbols in the transmission data experience the same signal-to-noise ratio, and the energy of the pilot is The energy is allocated to the transmission data, and there are two parts of the pilot in the structure frame, which can be allocated to The energy of the pilot, so

[0134] Step 2: Use a pulse shaper with a sampling rate of T s to convert s[n] from a discrete signal to a baseband continuous signal where ρ(t) is the pulse shaper, and the bandwidth of the baseband signal is After the loading wave f c operation, the passband signal corresponding to the structure frame is expressed as

[0135] Step 3: The passband transmission signal is transmitted to the underwater environment by converting the electrical signal into an acoustic signal through a power amplifier, a matching circuit, and a transducer.

[0136] Step 4: The underwater communication environment simulated by the SSML channel model includes three interference factors: Doppler scaling factor, channel gain, and channel delay. The SSML channel model is expressed as where τ p and A p are the transmission delay and channel gain of the pth path, P is the number of multipath, and α is the Doppler scaling factor, which controls the compression and expansion scale of the signal. In addition, w(t) is the AWGN, which is subject to zero mean distribution, and the variance is σ 2 .

[0137] Step 5: The signal transmitted through the underwater acoustic channel is received by the hydrophone and converted into an electrical signal. The signal is enhanced after passing through the pre-amplification module, and the passband OCDM underwater acoustic communication system receiver signal after passing through the SSML channel is obtained through the band-pass filter

[0138] Step 6: oversample the received signal with a sampling rate of r , where the oversampling rate is After synchronization processing, the discrete passband received signal is denoted as A matched filter is used for the oversampled signal, taking the first chirp signal in the first block after synchronization as the template, and performing autocorrelation operation to generate multiple correlation peaks. The time of arrival of the ith correlation peak in the passband received signal is where the Doppler search subset is related to the maximum scale variation, containing [-α max (λL+λMi), α max (λL+λMi)], where α max is the maximum Doppler scale; the starting block index and the ending block index are determined. The Doppler scaling factor is calculated using adjacent correlation peaks, and if it exceeds a set threshold, it indicates that the variation is too large, i.e., the relative speed between the transmitter and the receiver has changed rapidly during the transmission of the signal, and the signal needs to be processed in segments. The threshold discrimination method is where Γ th is the threshold for determining whether the received signal needs to be processed in segments. If the difference between the Doppler scaling factors of two adjacent blocks is greater than the threshold, the received signal is processed in segments, and according to the block index i, the starting block index K1 and the ending block index K2 in each segment are set.

[0139] Step 7: between the index blocks K1 and K2, the following operations are performed. For the OCDM discrete passband received signal, a multi-peak Doppler estimation algorithm is used to reasonably utilize the multiple peaks to estimate the Doppler scaling factor, which is specifically expressed as The Doppler scaling factor estimated by the multi-peak Doppler estimation The resampling operation is performed on the OCDM passband received signal, and the resampled OCDM received signal is denoted as The down-sampling operation is performed on the resampled signal with a down-sampling ratio of , and the carrier f c is removed. The baseband signal can be represented as where LPF{·} is a low-pass filter, and the OCDM baseband discrete signal can be represented as Assuming that 1+α and are very close, i.e., the estimated Doppler scaling factor used in the resampling operation has high accuracy, we can represent the OCDM baseband signal as where​ It's the CFO, w o ∈(0,1).

[0140] Step 8: After performing a serial-to-parallel conversion operation on the resampled signal, express y[n] in matrix form, where The i-th signal block can be expressed as The CFO parameter is determined by the first chirp signal [y(i)] in the i-th block. 1:L and the next in the preceding receive block a symbol To make an estimate, among which L h The maximum channel delay is given. Two sequences are concatenated to obtain a new sequence for estimating the CFO within the i-th block. The new sequence is represented as... The product of these symbol pairs is The closed-form expression for the CFO of the i-th block is: For the resampled signal, use the values ​​calculated using the closed-loop CFO described above. To compensate for the CFO value within the signal If the CFO is perfectly compensated

[0141] Step 9: Estimate the channel parameters using the first chirped signal in the i-th block. The MMSE channel estimator is... ZF channel estimator is During preprocessing, the trailing interference of preceding adjacent chirped signals is first removed from the OCDM signal block, represented as... in It is an L×L cyclic matrix, with the first column being [h(0),…,h(L)]. h ),0,…,0] T Next, the tail of the OCDM signal block is added to the received OCDM signal block through a cyclic superposition operation, and then the matrix F is left-multiplied. N The signal is shifted to the frequency domain, and the preprocessed signal expression is: Equalization is performed on the preprocessed signal, and the estimated OCDM data symbols of the received signal are represented as follows: ZF equalizer is G ZF (i)=Λ -1 (i) The MMSE equalizer is

[0142] like Figure 1As shown, in actual use, steps 1 to 3 are operations of the sending end, and steps 5 to 9 are operations of the receiving end; in implementation, the sending information is fixed after being executed once at the sending end, and is repeatedly sent at different times and different positions, and steps 5 to 9 are repeatedly executed.

[0143] The preferred embodiments of the present application are described in detail above. It should be understood that those of ordinary skill in the art can make modifications and variations without creative effort based on the concept of the present application. Therefore, any technical solution obtained by logical analysis, reasoning or limited experiments based on the concept of the present application and the prior art should be within the protection scope defined by the claims.

Claims

1. A multi-peak Doppler estimation compensation method based on an OCDM underwater acoustic communication system, characterized in that, Includes the following steps: Step 1: Generate a transmission frame in the transmitter of the OCDM underwater acoustic communication system, containing a length of... pilot symbols , length is Data symbols ,in The Fresnel inverse transformation matrix is... The data to be transmitted is combined into a format of length [length missing]. OCDM signal block , co-generated Each OCDM signal block, after serial-to-parallel conversion, has preambles and pre- and post-codes added. Energy and bandwidth are allocated separately. Step 2: Use a pulse shaper on the transmitted frame. Obtain continuous passband signal ; Step 3: Convert the continuous passband signal The electrical signal is converted into an acoustic signal and transmitted to the underwater environment through a power amplifier, matching circuit and transducer; Step 4: Use the SSML channel model To simulate an underwater communication environment; Step 5: Represents the received signal after the OCDM underwater acoustic communication device transmits the signal through the SSML channel. ; Step 6: Oversample the received signal at an oversampling rate. Perform an oversampling operation to obtain a discrete passband received signal. ,right Perform autocorrelation operation, the template is the length after the received signal synchronization point. sequence, The search range within each block and related to the Doppler scaling factor is used to obtain multiple Doppler-related peaks. Then, a threshold discrimination method was used. The Doppler scaling factor calculated for each single peak is: , To determine the threshold; The threshold method determines whether the relative speeds of the transmitter and receiver change significantly during signal transmission. If so, segmented processing is required to determine the starting block index. and end block index If not, then set ; Step 7, in The Doppler scaling factor is estimated within the interval using the multi-peak Doppler estimation method. For discrete passband received signals Perform a resampling operation to initially compensate for the Doppler scaling factor and obtain the resampled OCDM received signal. Then, a downsampling and carrier removal operation is performed to obtain the OCDM baseband discrete signal. The signal contains residual Doppler. The impact; Step 8: Resample the signal Perform a serial-to-parallel conversion operation to obtain the first... Signal block expression Using the newly spliced ​​sequence Estimate the first The CFO parameters in each block, among which Given the uninterrupted symbol length, write the closed-form expression for CFO in the signal. And compensate the signal block to obtain the compensated signal. ; Step 9: Removing Doppler effects from the signal Perform channel estimation operations to obtain a channel estimator and ; Perform preprocessing operations, including removing the tails of adjacent sequences and cyclically superimposing its own tails, to obtain a preprocessed signal. Perform equalization operation to obtain the OCDM symbols estimated from the received signal. .

2. The multi-peak Doppler estimation compensation method for an OCDM-based underwater acoustic communication system as described in claim 1, characterized in that, In step 1, the OCDM signal block is modulated using a discrete Fresnel matrix, the size of which is... The DFnT matrix Input is ,in The number of subcarriers in the OCDM system; the OCDM structure frame consists of payload information. and two lengths The known sequence composition and ; length is The data symbol vector is taken from a complex modulation alphabet and represented as follows: ;in and Defined as Pilot symbols are designed as Fixed power They are assigned to two known sequences respectively; the first sequence in the transmitted frame The length of each transmitted signal block is , written as ; Pilot signals need to cover the entire bandwidth. , forming a transport block The frequency resolutions of the three parts are respectively and ; After the serial-to-parallel conversion, add a length of [length missing] at the beginning and end. known sequence These two known sequences are also defined as The first generation of baseband transmission signal Input Specifically, Utilizing energy Transmit OCDM structure frames, containing One transmission block; ensuring that symbols in both the pilot and transmitted data experience the same signal-to-noise ratio, Energy is allocated to transmitted data; the structured frame contains two pilot components, each of which can be allocated to... The energy, therefore .

3. The multi-peak Doppler estimation compensation method for an OCDM underwater acoustic communication system as described in claim 1, characterized in that, Step 6 includes the following steps: Step 6.1: Perform an oversampling operation on the passband OCDM received signal, with an oversampling rate of... ,in The sampling frequency for the transmitted signal. To obtain the discrete passband received signal, the sampling frequency of the received signal is determined. ; Step 6.2: Receive discrete passband signals Perform an autocorrelation operation to obtain the first chirp signal in the first block after synchronization. Using this as a template, multiple Doppler correlation peaks were obtained. Doppler subset search Related to the largest scale change ,in It is the largest Doppler scaling scale; Step 6.3: Use the threshold discrimination method. The Doppler scaling factor calculated by single-peak calculation , It is the threshold for determining whether the received signal needs to be segmented; it determines the starting block index for subsequent operations. and end block index .

4. The multi-peak Doppler estimation compensation method for an OCDM-based underwater acoustic communication system as described in claim 1, characterized in that, Step 7 includes the following steps: Step 7.1, in the index block and The following operations are performed between these operations, using multiple Doppler correlation peaks for the OCDM discrete passband received signal. Estimate Doppler scaling factor ,in represent and the Minimum interval block index between relevant peaks; Step 7.2: Utilize the estimated Doppler scaling factor For OCDM passband received signals Perform a resampling operation to obtain the resampled OCDM received signal. ; Step 7.3: Resample the signal by The downsampling ratio performs downsampling and carrier removal operations. ,in For carrier frequency, The sampling frequency for the transmitted signal. To obtain the OCDM baseband received signal, the sampling frequency is set. ; Step 7.4, Assumption and When they are very close, the OCDM baseband signal after the resampling operation is re-represented. Among them, residual Doppler , , , , It is a baseband AWGN.

5. The multi-peak Doppler estimation compensation method for an OCDM-based underwater acoustic communication system as described in claim 1, characterized in that, Step 8 includes the following steps: Step 8.1: Perform a serial-to-parallel conversion operation on the resampled signal to convert the OCDM baseband signal. Expressed in matrix form, where ; obtained the first Signal block expression ,in Residual Doppler interference within the expression block The channel parameter matrix, It is the inverse Fresnel transformation matrix; Step 8.2: Use the newly spliced ​​sequence Estimate the first The CFO parameters in each block, and the closed-form expression for CFO interference within the signal. ; Step 8.3: For the resampled signal, use the estimated CFO value. The compensation signal has a CFO (Cost Forward Fault) if the CFO in the signal is completely eliminated and the subcarriers in the compensation signal are orthogonal, thus obtaining the compensation signal. .

6. The multi-peak Doppler estimation compensation method for an OCDM underwater acoustic communication system as described in claim 1, characterized in that, Step 9 includes the following steps: Step 9.1: Perform channel estimation on the signal after removing Doppler effects to obtain the channel estimator. ,in, For channel variance, For noise variance, and These are the pilot signal values ​​and the allocated power, respectively; when the channel variance is unavailable, use... ; Step 9.2: Removing Doppler effects from the signal Preprocessing is performed, firstly removing the trailing interference of preceding adjacent chirped signals from the OCDM signal block, represented as... Next, the tail of the OCDM signal block is added to the received OCDM signal block through a cyclic superposition operation, and then the Fourier transform matrix is ​​left-multiplied. The signal is shifted to the frequency domain, and the preprocessed signal expression is: ; Step 9.3: Process the preprocessed signal Perform an equalization operation to obtain the estimated OCDM data symbols. ,in The diagonalized Fresnel transform parameter matrix is ​​used for the ZF equalizer. , where the diagonal matrix It is the channel frequency response. It is a cyclic superposition matrix, and the MMSE equalizer is... ,in Let be the power of the transmitted data symbols in the structured frame, and let the additive noise variance matrix be . ,in This represents the noise variance.

7. The multi-peak Doppler estimation and compensation method for an OCDM-based underwater acoustic communication system as described in claim 1, characterized in that, The underwater acoustic communication transmitter includes a digital-to-analog converter module, a power amplifier, a matching circuit, and a transducer.

8. The multi-peak Doppler estimation compensation method for an OCDM underwater acoustic communication system as described in claim 1, characterized in that, The underwater acoustic communication receiver includes an acoustic-to-electric conversion module, a signal amplification module, and a bandpass filter.

9. The multi-peak Doppler estimation compensation method for an OCDM-based underwater acoustic communication system as described in claim 1, characterized in that, Steps 1 to 3 are operations performed on the transmitting end of the underwater acoustic communication device.

10. The multi-peak Doppler estimation compensation method for an OCDM-based underwater acoustic communication system as described in claim 1, characterized in that, Steps 5 to 9 are the operations of the underwater acoustic communication receiver.

Images