A deep-sea large-delay sparse cluster diversity MIMO underwater acoustic communication channel equalization method

Abstract

The application belongs to the technical field of underwater acoustic communication and particularly relates to a deep-sea large-time-delay sparse cluster diversity MIMO underwater acoustic communication channel equalization method, which comprises the following steps: establishing a multiple-input multiple-output (MIMO) underwater acoustic communication signal receiving model under a deep-sea large-time-delay channel, extracting the arrival time delay of channel tap coefficients by using matched filtering, decomposing the channel into multiple sub-channels, and constructing a receiving signal vector of each sub-channel; based on the receiving signal vector of each sub-channel, constructing a clustered receiving signal vector and a clustered interference signal vector, and completing large-time-delay time-varying multipath interference elimination through adaptive turbo equalization, thereby solving the problem of deep-sea large-time-delay underwater multipath channel interference suppression under the condition of a traditional short training sequence.

Description

Technical Field

[0001] This invention belongs to the field of underwater acoustic communication technology, specifically relating to a channel equalization method for deep-sea long-delay sparse cluster diversity MIMO underwater acoustic communication. Background Technology

[0002] With the deepening of marine resource exploration and development, the demand for information transmission in deep-sea environments is becoming increasingly urgent. Underwater acoustic communication, as the only reliable long-distance communication method underwater, directly affects the safety and efficiency of deep-sea operations. However, deep-sea acoustic channels have significant unique characteristics: complex sea conditions such as thermoclines and internal waves lead to highly random signal propagation paths, resulting in large time delays, reaching hundreds or even thousands of milliseconds in a sparse multipath structure; limited bandwidth and high propagation loss exacerbate signal distortion; simultaneously, Doppler shift caused by platform motion further deteriorates the quality of the received signal. These factors collectively lead to severe inter-symbol interference at the receiving end, becoming a core bottleneck restricting communication speed and reliability.

[0003] Traditional underwater acoustic communication often employs a single-input single-output (SISO) architecture, using adaptive equalizers or channel estimation-based equalizers to suppress inter-symbol interference (ISI). However, given the large time delays and sparse distribution of multipath clusters in the deep sea, traditional equalizers require an extremely large number of taps to cover the delay spread, resulting in exponentially increasing computational complexity. Furthermore, their ability to track dynamic channels is limited, easily leading to increased bit error rates. In recent years, multiple-input multiple-output (MIMO) technology has been introduced into underwater acoustic communication due to its ability to improve communication capacity through spatial diversity and multiplexing. However, deep-sea MIMO channels not only inherit the delay-Doppler dual distortion of SISO channels but also suffer from spatial correlation between multiple antennas and cluster diversity characteristics—that is, multipath energy is concentrated in a few discrete clusters. This makes it difficult for traditional MIMO equalization methods to effectively separate overlapping substreams, especially with a sharp decline in performance at low signal-to-noise ratios.

[0004] Furthermore, while existing diversity techniques such as time / frequency diversity can enhance fading resistance, they are not optimized for the non-uniform energy distribution of sparse cluster diversity channels in the deep sea, resulting in limited diversity gain. Therefore, there is an urgent need for an efficient equalization method that can adapt to the characteristics of long latency and sparse cluster diversity MIMO channels in the deep sea, reducing computational complexity while improving multipath energy utilization and dynamic channel tracking capabilities to meet the requirements of high-speed and reliable communication in the deep sea.

[0005] Current research largely focuses on shallow-sea channels, lacking specific design for MIMO equalization in extreme deep-sea environments, particularly in terms of modeling and utilizing coefficient cluster diversity structures. This invention addresses these issues by proposing a deep-sea MIMO channel processing method that integrates diversity gain and adaptive equalization. This aims to overcome existing technological bottlenecks and provide theoretical support and engineering implementation solutions for deep-sea communication. Summary of the Invention

[0006] This invention provides the following technical solution:

[0007] A channel equalization method for deep-sea long-delay sparse cluster diversity MIMO underwater acoustic communication, the method comprising the following steps:

[0008] Step 1) Establish a multi-input multi-output (MIMO) underwater acoustic communication signal reception model under a deep-sea long-delay channel;

[0009] Step 2) Use matched filtering to extract the arrival delay of the channel tap coefficients, decompose the channel into multiple sub-channels, and construct the received signal vector of each sub-channel;

[0010] Step 3) Based on the received signal vector of each sub-channel Construct clustered received signal vectors and clustered interference signal vector Adaptive turbo equalization is used to eliminate time-varying multipath interference with large time delays.

[0011] Furthermore, step 1) specifically includes:

[0012] For a multiple-input multiple-output single-carrier underwater acoustic communication system, the number of transmitting transducers is: The number of receiving hydrophones is The received signal model is represented as

[0013]

[0014] in, Indicates the first Each receiving element in The signal value received at any time, Indicates in Time of the first The first transmitting transducer and the first The first impulse response of a deep-sea long-delay channel between hydrophones One tap, channel impulse response length is , express Time of the first Transmitted symbols from each transmitter transducer Indicates the first Each receiving element in Environmental noise at all times;

[0015] definition for time The received signal vector of each receiving array element This indicates the transpose operation.

[0016]

[0017] in, for The first moment of the MIMO channel A tap coefficient matrix for time The transmitted symbol vector of each transmitting transducer for time The noise vector of each receiving array element is then further expressed as:

[0018]

[0019] Among them, large-delay cluster sparse channels delay spread length ,contain A multipath cluster, the first The starting tap index of each cluster head is The intra-cluster multipath delay spread length is The channel tap coefficient between adjacent clusters is 0, and the interval between adjacent clusters is defined as follows:

[0020]

[0021] The model is now complete.

[0022] Furthermore, step 2) specifically includes:

[0023] Step 2-1) Demodulate the received signal to baseband, and use the classic synchronization header detection method based on matched filtering to complete data frame synchronization. Based on the matched filtering result, obtain... One transmitting transducer arrived Arrival delay of channel tap coefficients for each receiving array element;

[0024] Step 2-2) Based on the multipath spread characteristics of deep-sea long-delay channels, the channel is decomposed in the delay dimension into... The nth channel multipath cluster is defined as the nth... The starting tap index of each channel cluster is The starting tap index of the first channel cluster is extracted as follows: , No. The starting tap index of each channel cluster Determined as The index of the first channel tap with a matched filter amplitude greater than 0.2;

[0025] Steps 2-3) Treat each channel cluster as a sub-channel, construct the received signal of each sub-channel, and define it in... Time of the first The received signal vectors of each sub-channel are represented as follows: .

[0026] Furthermore, step 3) specifically includes:

[0027] Step 3-1) Construct for the first The first of the launch array elements Clustered received signal vectors with symbol equalization ;

[0028] Step 3-2) Construct for the first The first of the launch array elements A symbol-equalized clustered interference signal vector ;

[0029] Step 3-3) Based on the constructed clustered received signal vector and clustered interference signal vector Perform adaptive turbo balancing.

[0030] Furthermore, step 3-1) specifically includes:

[0031] Step 3-1-1) Construct the first Received signal vectors of each channel cluster , length is ;

[0032] Step 3-1-2) Construct clustered received signal vectors .

[0033] Furthermore, step 3-2) specifically includes:

[0034] Step 3-2-1) Construction The disturbance estimation vector at time t.

[0035]

[0036] in This indicates the result of conventional computation based on the decoder output soft information obtained in this turbo iteration. The first of the launch array elements Prior estimates of a symbol;

[0037] Step 3-2-2) Based on the interference estimation vector , No. The first of the launch array elements The first symbolic equalization required by the symbolic equalization Channel cluster interference signal vector The specific construction method is as follows

[0038] ;

[0039] Step 3-2-3) Construct for the first The first of the launch array elements A symbol-equalized clustered interference signal vector , , length is .

[0040] Furthermore, step 3-3) specifically includes:

[0041] Step 3-3-1) For the first The first of the launch array elements symbols, constructing a length of feedforward filter coefficient vector , length is Interference cancellation filter coefficient vector ,in The value should be set according to the actual underwater acoustic channel environment. and Set as a vector of all zeros;

[0042] Step 3-3-2) For the first The first of the launch array elements Equilibrium results of symbols Solution: ,in This represents the conjugate transpose operation;

[0043] Step 3-3-3) Update the feedforward filter coefficients symbol by symbol using the commonly used normalized least mean square adaptive filtering algorithm or the improved proportionally normalized least mean square adaptive filtering algorithm. Quantity and interference cancellation filter coefficient vector Simultaneously obtain the equilibrium result of all symbols in each transmitting element;

[0044] Step 3-3-4) Using a general soft decision and demapping calculation method, the equalization results of all symbols are converted into bit log-likelihood ratio soft information and input into the soft decoder for the next turbo iteration.

[0045] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects:

[0046] This invention reduces the overhead requirements of channel equalization on the training sequence by transforming the equalization of large-delay clusters of sparse channels into joint equalization of multiple short-delay sub-channels. Unlike classical equalization processes where the training sequence length is much longer than the channel delay spread, this invention only requires the training sequence length to be greater than the multipath delay spread of the sub-channels to ensure convergence of the equalization taps of each sub-channel. Simultaneously, it utilizes the differences between different channel clusters to achieve diversity gain. When the delay interval between adjacent clusters is greater than the data frame length, in the time domain, it is equivalent to receiving multiple communication signals that have undergone transmission through different channels without interference. In this case, the method can obtain the maximum combining gain. Attached Figure Description

[0047] Figure 1 This is a flowchart of a deep-sea long-delay sparse cluster diversity MIMO underwater acoustic communication channel equalization method according to the present invention;

[0048] Figure 2 This is a graph showing the change of the impulse response of the deep-sea acoustic communication channel over time as measured by this invention.

[0049] Figure 3 It is a constellation diagram output after being processed by the method described in this invention. Detailed Implementation

[0050] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings, so that those skilled in the art can more clearly understand how to practice the present invention. Although the present invention has been described in conjunction with its preferred embodiments, these embodiments are merely illustrative and not intended to limit the scope of the invention.

[0051] As attached Figure 1-3 As shown, a channel equalization method for deep-sea long-delay sparse cluster diversity MIMO underwater acoustic communication includes the following steps:

[0052] Step 1) Establish a multi-input multi-output (MIMO) underwater acoustic communication signal reception model under a deep-sea long-delay channel;

[0053] Step 2) Use matched filtering to extract the arrival delay of the channel tap coefficients, decompose the channel into multiple sub-channels, and construct the received signal vector of each sub-channel;

[0054] Step 3) Based on the received signal vector of each sub-channel Construct clustered received signal vectors and clustered interference signal vector The large-delay time-varying multipath interference is eliminated through adaptive turbo equalization.

[0055] Specifically, step 1) is as follows:

[0056] For a multiple-input multiple-output single-carrier underwater acoustic communication system, the number of transmitting transducers is: The number of receiving hydrophones is The received signal model is represented as

[0057]

[0058] in, Indicates the first Each receiving element in The signal value received at any time, Indicates in Time of the first The first transmitting transducer and the first The first impulse response of a deep-sea long-delay channel between hydrophones One tap, channel impulse response length is , express Time of the first Transmitted symbols from each transmitter transducer Indicates the first Each receiving element in Environmental noise at all times;

[0059] definition for time The received signal vector of each receiving array element This indicates the transpose operation.

[0060]

[0061] in, for The first moment of the MIMO channel A tap coefficient matrix for time The transmitted symbol vector of each transmitting transducer for time The noise vector of each receiving array element is then further expressed as:

[0062]

[0063] Among them, large-delay cluster sparse channels delay spread length ,contain A multipath cluster, the first The starting tap index of each cluster head is The intra-cluster multipath delay spread length is The channel tap coefficient between adjacent clusters is 0, and the interval between adjacent clusters is defined as follows:

[0064]

[0065] The model is now complete.

[0066] Specifically, step 2) includes:

[0067] Step 2-1) Demodulate the received signal to baseband, and use the classic synchronization header detection method based on matched filtering to complete data frame synchronization. Based on the matched filtering result, obtain... One transmitting transducer arrived Arrival delay of channel tap coefficients for each receiving array element;

[0068] Step 2-2) Based on the multipath spread characteristics of deep-sea long-delay channels, the channel is decomposed in the delay dimension into... The nth channel multipath cluster is defined as the nth... The starting tap index of each channel cluster is The starting tap index of the first channel cluster is extracted as follows: , No. The starting tap index of each channel cluster Determined as The index of the first channel tap with a matched filter amplitude greater than 0.2;

[0069] Steps 2-3) Treat each channel cluster as a sub-channel, construct the received signal of each sub-channel, and define it in... Time of the first The received signal vectors of each sub-channel are represented as follows: .

[0070] Specifically, step 3) includes:

[0071] Step 3-1) Construct for the first The first of the launch array elements Clustered received signal vectors with symbol equalization ;

[0072] Step 3-2) Construct for the first The first of the launch array elements A symbol-equalized clustered interference signal vector ;

[0073] Step 3-3) Based on the constructed clustered received signal vector and clustered interference signal vector Perform adaptive turbo balancing.

[0074] Specifically, step 3-1) includes:

[0075] Step 3-1-1) Construct the first Received signal vectors of each channel cluster , length is ;

[0076] Step 3-1-2) Construct clustered received signal vectors .

[0077] Specifically, step 3-2) includes:

[0078] Step 3-2-1) Construction The disturbance estimation vector at time t.

[0079]

[0080] in This indicates the result of conventional computation based on the decoder output soft information obtained in this turbo iteration. The first of the launch array elements Prior estimates of a symbol;

[0081] Step 3-2-2) Based on the interference estimation vector , No. The first of the launch array elements The first symbolic equalization required by the symbolic equalization Channel cluster interference signal vector The specific construction method is as follows

[0082] ;

[0083] Step 3-2-3) Construct for the first The first of the launch array elements A symbol-equalized clustered interference signal vector , , length is .

[0084] Specifically, step 3-3) includes:

[0085] Step 3-3-1) For the first The first of the launch array elements symbols, constructing a length of feedforward filter coefficient vector , length is Interference cancellation filter coefficient vector ,in The value should be set according to the actual underwater acoustic channel environment. and Set as a vector of all zeros;

[0086] Step 3-3-2) For the first The first of the launch array elements Equilibrium results of symbols Solution: ,in This represents the conjugate transpose operation;

[0087] Step 3-3-3) Update the feedforward filter coefficients symbol by symbol using the commonly used normalized least mean square adaptive filtering algorithm or the improved proportionally normalized least mean square adaptive filtering algorithm. Quantity and interference cancellation filter coefficient vector Simultaneously, the equilibrium result of all symbols in each transmitting element is obtained;

[0088] Step 3-3-4) Using a general soft decision and demapping calculation method, the equalization results of all symbols are converted into bit log-likelihood ratio soft information and input into the soft decoder for the next turbo iteration.

[0089] In addition, step 3) can be implemented by various existing channel equalization methods based on channel estimation, such as linear MMSE equalizers and decision feedback equalizers. However, compared with these methods, the present invention can track channel changes symbol by symbol, has lower computational complexity, and stronger resistance to time-varying conditions.

[0090] The method described in this invention processes underwater acoustic communication waveforms acquired in deep-sea experiments. The experiments employed BPSK modulation with a communication bandwidth of 2kHz, a training sequence length of 1000, and a corresponding duration of 50ms. (See attached diagram) Figure 2 The paper presents the measured impulse response of a deep-sea acoustic communication channel over time. The channel spread delay is approximately 120 ms, comprising three channel clusters located near the multipath delay positions of 0 ms, 25 ms, and 120 ms. This channel spread delay is significantly longer than the system training sequence duration, rendering traditional communication methods ineffective for accurate reception and decoding. (See attached image) Figure 3 The diagram shown is a constellation image output after processing by the method proposed in this invention. (See attached image.) Figure 3 As can be seen, the constellation diagram performance gradually improves with the increase of the number of iterations, and the method has significant performance advantages, verifying the effectiveness of the deep-sea long-delay sparse cluster diversity MIMO underwater acoustic communication channel equalization method proposed in this invention.

[0091] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications and substitutions based on the technical solutions and inventive concepts provided by the present invention should be covered within the scope of protection of the present invention.

Claims

1. A channel equalization method for deep-sea long-delay sparse cluster diversity MIMO underwater acoustic communication, characterized in that, The method includes the following steps: Step 1) Establish a multi-input multi-output (MIMO) underwater acoustic communication signal reception model under a deep-sea long-delay channel; Step 2) Use matched filtering to extract the arrival delay of the channel tap coefficients, decompose the channel into multiple sub-channels, and construct the received signal vector of each sub-channel; Step 3) Based on the received signal vector of each sub-channel Construct clustered received signal vectors and clustered interference signal vector The large-delay time-varying multipath interference is eliminated through adaptive turbo equalization.

2. The channel equalization method for deep-sea long-delay sparse cluster diversity MIMO underwater acoustic communication according to claim 1, characterized in that, Step 1) specifically refers to: For a multiple-input multiple-output single-carrier underwater acoustic communication system, the number of transmitting transducers is: The number of receiving hydrophones is The received signal model is represented as in, Indicates the first Each receiving element in The signal value received at any time, Indicates in Time of the first The first transmitting transducer and the first The first impulse response of a deep-sea long-delay channel between hydrophones One tap, channel impulse response length is , express Time of the first Transmitted symbols from each transmitter transducer Indicates the first Each receiving element in Environmental noise at all times; definition for time The received signal vector of each receiving array element This indicates the transpose operation. in, for The first moment of the MIMO channel A tap coefficient matrix for time The transmitted symbol vector of each transmitting transducer for time The noise vector of each receiving array element is then further expressed as: Among them, large-delay cluster sparse channels delay spread length ,contain A multipath cluster, the first The starting tap index of each cluster head is The intra-cluster multipath delay spread length is The channel tap coefficient between adjacent clusters is 0, and the interval between adjacent clusters is defined as follows: The model is now complete.

3. The channel equalization method for deep-sea long-delay sparse cluster diversity MIMO underwater acoustic communication according to claim 1, characterized in that, Step 2) specifically includes: Step 2-1) Demodulate the received signal to baseband, and use the classic synchronization header detection method based on matched filtering to complete data frame synchronization. Based on the matched filtering result, obtain... One transmitting transducer arrived Arrival delay of channel tap coefficients for each receiving array element; Step 2-2) Based on the multipath spread characteristics of deep-sea long-delay channels, the channel is decomposed in the delay dimension into... The nth channel multipath cluster is defined as the nth... The starting tap index of each channel cluster is The starting tap index of the first channel cluster is extracted as follows: , No. The starting tap index of each channel cluster Determined as The index of the first channel tap with a matched filter amplitude greater than 0.2; Steps 2-3) Treat each channel cluster as a sub-channel, construct the received signal of each sub-channel, and define it in... Time of the first The received signal vectors of each sub-channel are represented as follows: .

4. The channel equalization method for deep-sea long-delay sparse cluster diversity MIMO underwater acoustic communication according to claim 1, characterized in that, Step 3) specifically includes: Step 3-1) Construct for the first The first of the launch array elements Clustered received signal vectors with symbol equalization ; Step 3-2) Construct for the first The first of the launch array elements A symbol-equalized clustered interference signal vector ; Step 3-3) Based on the constructed clustered received signal vector and clustered interference signal vector Perform adaptive turbo balancing.

5. The channel equalization method for deep-sea long-delay sparse cluster diversity MIMO underwater acoustic communication according to claim 4, characterized in that, Step 3-1) specifically includes: Step 3-1-1) Construct the first Received signal vectors of each channel cluster , length is ; Step 3-1-2) Construct clustered received signal vectors .

6. The channel equalization method for deep-sea long-delay sparse cluster diversity MIMO underwater acoustic communication according to claim 4, characterized in that, Step 3-2) specifically includes: Step 3-2-1) Construction The disturbance estimation vector at time t. in This indicates the result of conventional computation based on the decoder output soft information obtained in this turbo iteration. The first of the launch array elements Prior estimates of a symbol; Step 3-2-2) Based on the interference estimation vector , No. The first of the launch array elements The first symbolic equalization required by the symbolic equalization Channel cluster interference signal vector The specific construction method is as follows ； Step 3-2-3) Construct for the first The first of the launch array elements A symbol-equalized clustered interference signal vector , , length is .

7. A channel equalization method for deep-sea long-delay sparse cluster diversity MIMO underwater acoustic communication according to claim 4, characterized in that, Step 3-3) specifically includes: Step 3-3-1) For the first The first of the launch array elements symbols, constructing a length of feedforward filter coefficient vector , length is Interference cancellation filter coefficient vector ,in The value should be set according to the actual underwater acoustic channel environment. and Set as a vector of all zeros; Step 3-3-2) For the first The first of the launch array elements Equilibrium results of symbols Solution: ,in This represents the conjugate transpose operation; Step 3-3-3) Update the feedforward filter coefficients symbol by symbol using the commonly used normalized least mean square adaptive filtering algorithm or the improved proportionally normalized least mean square adaptive filtering algorithm. Quantity and interference cancellation filter coefficient vector Simultaneously obtain the equilibrium result of all symbols in each transmitting element; Step 3-3-4) Using a general soft decision and demapping calculation method, the equalization results of all symbols are converted into bit log-likelihood ratio soft information and input into the soft decoder for the next turbo iteration.

Images