Training sequence updating method for underwater acoustic communication channel estimation
By jointly designing channel encoding and signal frames in water acoustic communication, and updating training sequences using error correction decoding and CRC verification, the problems of channel estimation obsolete and communication rate reduction are solved, and more efficient water acoustic communication performance is achieved.
Patent Information
- Application Number
- CN202510377824.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-28
- Publication Date
- 2025-06-20
AI Technical Summary
In water acoustic communication, the multipath expansion and fast time-varying characteristics of the channel cause the channel estimation to be outdated, which in turn affects the communication performance. The frequent insertion of training sequences increases communication overhead and reduces the rate.
By performing a joint design of channel encoding and signal frames on the transmitting end, the training sequence is updated using error correction decoding and CRC verification results on the receiving end, avoiding frequent insertion of training sequences, and dynamic update of channel estimation is realized.
The problems of channel estimation outdated and communication rate reduction are effectively avoided, the performance of water acoustic communication in fast time-varying channels is improved, and the accuracy of channel estimation and communication efficiency are ensured.
Smart Images
Figure CN120185773A_ABST
Abstract
Description
Technical Field:
[0001] The present invention belongs to the technical field of underwater acoustic communication, and particularly relates to a training sequence update method for underwater acoustic communication channel estimation. Background Art:
[0002] Underwater acoustic channels have severe multipath spread and fast time-varying characteristics, which are the main factors restricting the improvement of underwater acoustic communication performance. Accurate underwater acoustic channel estimation is the key to improving the decoding accuracy rate. Especially for high-speed coherent underwater acoustic communication technology, the channel estimation performance is directly related to the channel equalization effect, and thus affects the bit error rate of underwater acoustic communication. The requirements of underwater acoustic communication for channel estimation are: first, being able to accurately estimate the intensity and phase of the multipath tap coefficients of the channel transfer function; second, being able to track the fast changes of the channel. Underwater acoustic communication usually uses an adaptive technique for the training sequence to perform channel estimation. The length of the training sequence is one of the elements determining the quality of channel estimation performance. The longer the training sequence length, the better the estimation performance. Generally, a training sequence several times the length of the maximum multipath spread of the channel is inserted at the coding signal head of the communication signal frame to obtain an accurate estimation of the channel transfer function. Considering the fast time-varying nature of the underwater acoustic channel, to prevent the channel transfer function of the coding signal far from the end of the training sequence frame from becoming outdated, it is necessary to frequently insert the training sequence, which will result in a large overhead of underwater acoustic communication, thereby greatly reducing the rate of underwater acoustic communication. Summary of the Invention:
[0003] The technical problem to be solved by the present invention is to provide a training sequence update method for underwater acoustic communication channel estimation. This method performs joint design on channel coding and signal frames at the sending end, performs error correction decoding and CRC check processing at the receiving end, and updates the training sequence according to the CRC check result, without the need to frequently insert the training sequence to track the fast time-varying channel, avoiding the problems of reduced communication rate and outdated channel estimation. This method is simple to implement and has good practical value.
[0004] The technical solution of the present invention is to provide a training sequence update method for underwater acoustic communication channel estimation, including the following steps,
[0005] Channel coding: At the sending end, map the bit stream to be transmitted into code elements and divide them into N groups on average. Each group is divided into M blocks on average. First, perform CRC coding on each block of code elements, then perform error correction coding on the entire group of processed code elements, and then divide the error correction coding check code elements into N blocks on average and disperse them into each block to form new blocks. Finally, perform interleaving on the information code elements, CRC check code elements, and error correction code check code elements of each new block to complete the channel coding process;
[0006] Signal frame design
[0007] Take the first block of code elements in each group after channel coding and perform symbol mapping and signal modulation in sequence to form N coded signals for the first half frame. Take the second block of code elements and perform signal modulation in sequence to form N coded signals for the second half frame. Insert a training sequence signal between the first and second half frames to form the first frame of communication signal. Take the coded signal generated from the third block of code elements in each group as the first half frame, and the coded signal generated from the fourth block of code elements as the second half frame. Insert a training sequence signal between the first and second half frames to form the second frame of communication signal. And so on, finally a total of M / 2 frames of signals are formed, and each frame contains 2N coded signals. Since the present invention inserts the training sequence signal in the middle of the frame, so that the receiving end can perform channel equalization forward and backward using the channel estimation result. Compared with the traditional frame structure that only performs equalization backward, to a certain extent, it avoids the problem that the channel estimation result becomes outdated due to channel time variation.
[0008] Pilot sequence update. At the receiving end, perform channel estimation on the training sequence signal of each frame of the M / 2 frames of signals simultaneously to obtain M / 2 channel transfer functions. Take one block of coded signal before and after the training sequence signal of each frame, and perform channel equalization, demodulation, and deinterleaving processing in sequence using the channel transfer function of their respective frames. Then perform error correction decoding on the results of the M blocks after deinterleaving processing uniformly, and then perform CRC check on each block of decoding results. The block information code elements and CRC check code elements that pass the check are interleaved as the new training sequence to achieve training sequence update. And so on, use the new training sequence to perform channel estimation to obtain the channel transfer function, and take one block of coded signal forward and backward respectively to repeat the above process until all code blocks are decoded. During the entire decoding process, for the case of CRC check error, select the block that is closest in time to the signal block to be decoded and has a correct check as the training sequence to avoid the problem that the channel estimation performance deteriorates due to incorrect training sequence.
[0009] Compared with the prior art, the present invention has the following advantages:
[0010] The present invention proposes a method for updating the training sequence of underwater acoustic communication channel estimation, and its beneficial effects are as follows: ① This method updates the training sequence using the correctly received decoded information, without the need to frequently insert training sequences to track fast time-varying channels, avoiding the problems of reduced communication rate and outdated channel estimation; ② Through joint design of channel coding and signal frames, a combined interleaving method of intra-block interleaving and inter-block interleaving is formed, ensuring the error correction performance of error correction coding against burst errors. Since longer error correction code groups can be formed, it is more conducive to the exertion of error correction ability; ③ Different from the traditional underwater acoustic communication signal frame structure, the frame structure of the present invention makes the communication performance less affected by channel time variation. Description of the drawings:
[0011] Figure 1 It is the information group framing method of traditional underwater acoustic communication.
[0012] Figure 2 is the signal frame structure of traditional underwater acoustic communication.
[0013] Figure 3 is the information framing method of the underwater acoustic communication of the present invention.
[0014] Figure 4 is the signal frame structure of the underwater acoustic communication of the present invention.
[0015] Figure 5 is a schematic diagram of training sequence update. Specific implementation manner:
[0016] The present invention will be further described below in conjunction with the accompanying drawings with respect to the specific implementation manner:
[0017] Figure 1 shows a schematic diagram of the information framing method of traditional underwater acoustic communication. The information symbol a is successively subjected to CRC encoding, error correction encoding, and channel interleaving to generate symbol sequences b, c, and d, and then signal modulation is performed on d to generate an encoded signal s. Traditional underwater acoustic communication adopts a frame structure in which the encoded signal follows the training sequence signal. The receiving end uses the training sequence signal for underwater acoustic channel estimation, and then uses the channel estimation result for channel equalization. As Figure 2 shown, the problems existing in this frame structure are as follows: ① Due to the time-varying nature of the channel, the channel estimation result of the encoded signal far from the training sequence signal becomes outdated, thus affecting the channel equalization effect and easily causing receiving errors; ② Some high-speed coherent underwater acoustic communication technologies adopt the decision feedback equalization technology, and use the decision feedback result to update the channel estimation parameters again. When the channel multipath extension is serious, there is a problem that the equalizer does not converge due to incorrect decision feedback results, resulting in error propagation; ③ To overcome the influence of channel time-variation, it is necessary to frequently insert training sequence signals, which occupies a large communication overhead and causes a decrease in communication rate.
[0018] In view of the above problems, the present invention proposes a training sequence update method for underwater acoustic communication channel estimation. By jointly designing the channel coding and the signal frame (as Figure 3 , Figure 4 shown), the receiving end updates the training sequence in a forward and forward-backward sequential manner according to the parity check result through error correction decoding and CRC check (as Figure 5 shown). This method does not require frequent insertion of training sequence signals, and can largely avoid the problems of communication rate decrease and channel estimation obsolescence.
[0019] Figure 3 shows a schematic diagram of the information framing method of the present invention. At the sending end, the bit stream to be transmitted is mapped into symbols and then evenly divided into N groups. Each group is evenly divided into M blocks, a total of NM blocks. The code block of the nth group is represented as First, perform CRC encoding on each block (taking the nth group as an example) of code elements to obtain CRC check code elements. Then, for the entire group of code elements perform error correction encoding, and then evenly divide the error correction encoding check code element c (n) into M blocks and disperse them to each block to form new divided blocks. The mth code block can be expressed as Finally, for each (taking the mth block as an example) information code element of the new divided block and the CRC check code element the error correction code check code element perform interleaving respectively, and finally generate the encoding Complete the channel encoding process. Figure 4 The underwater acoustic communication signal frame structure of the present invention is given. Taking the first frame as an example, take the first block of code elements of each group after channel encoding and perform signal modulation in sequence to form the first half-frame encoded signal The second block of code elements perform signal modulation to form the second half-frame encoded signal Insert a training sequence signal between the first half-frame and the second half-frame to form the first frame of communication signal. Take the encoded signal generated by the third block of code elements of each group as the first half-frame, and the encoded signal generated by the fourth block of code elements as the second half-frame. Insert a training sequence signal between the first half-frame and the second half-frame to form the second frame of communication signal. And so on. Finally, a total of M / 2 frames of signals are formed, and each frame contains 2N blocks of encoded signals. The present invention performs in-block interleaving on each code block and the check code element After interleaving, the code blocks are dispersed to different signal frames, which can be regarded as inter-block interleaving. Therefore, the channel encoding and signal frames proposed by the present invention are jointly designed to form a combined interleaving method of in-block interleaving and inter-block interleaving, ensuring sufficient interleaving of the error correction encoding, and thus ensuring the error correction performance of the error correction encoding against burst errors. Generally speaking, the longer the error correction code, the better the error correction performance. The present invention can form a longer error correction code than the traditional underwater acoustic communication coding method, which is beneficial for the error correction code to exert a stronger error correction ability. Compare Figure 2 and Figure 4 , the frame structure design proposed by the present invention inserts the training sequence signal in the middle of the frame so that the receiving end can use the channel estimation result to perform channel equalization forward and backward. Compared with the traditional frame structure that only performs equalization backward, for transmitting the same number of bits of information, the encoding length of the present invention can be half of the traditional method, thus avoiding to a certain extent the problem that the channel estimation result becomes outdated due to channel time variation.
[0020] Figure 5The training sequence update process of the present invention is given. In the figure, the white blocks are the blocks decoded correctly, the gray ones represent the blocks decoded incorrectly, the solid arrows indicate the direction of updating the training sequence, and the dashed arrows indicate the direction where the update of the training sequence is prohibited. At the receiving end, first, channel estimation is performed using the training sequence signal of each frame of the M / 2 frame signal to obtain the channel transfer function. At the same time, one coding signal before and after the training sequence signal is taken respectively and subjected to channel equalization, demodulation, and deinterleaving processing in sequence. Then, the results of processing the M blocks are used as an error correction code group for error correction decoding. Next, CRC check is performed on the decoding result of each block. The block information bits and CRC check bits that pass the check are interleaved and used as a new training sequence to achieve the update of the training sequence. And so on, channel estimation is performed using the new training sequence to obtain the channel transfer function, and one coding signal is taken forward and backward respectively to repeat the above process until all code blocks are decoded. During the entire decoding process, in case of CRC check error, the block that is closest in time to the signal block to be decoded and passes the check is selected as the training sequence, avoiding the problem that the channel estimation performance deteriorates due to an incorrect training sequence.
[0021] The training sequence update method for underwater acoustic communication channel estimation proposed by the present invention forms an interleaving method that combines intra-block interleaving and inter-block interleaving through the joint design of channel coding and signal frames, ensuring the error correction ability of the error correction code against channel burst errors. Since longer error correction coding can be formed and it is more conducive to the exertion of the error correction ability, due to the frame structure design with the training sequence signal in the middle, compared with the traditional frame structure, the communication performance is less affected by channel time variation. The receiving end uses the error correction decoding and CRC check results to update the training sequence in a sequential manner forward and backward using the code blocks that pass the check, avoiding the problem that the channel estimation result becomes outdated due to channel time variation in the traditional underwater acoustic communication method, thereby effectively improving the communication performance of underwater acoustic communication in a fast time-varying channel. Since CRC check is adopted and only the code blocks that pass the check are selected as the training sequence, there is no problem that the channel estimation deteriorates due to the use of an incorrect training sequence.
[0022] The above is only an illustration of the preferred embodiments of the present invention and should not be construed as a limitation of the claims. All equivalent process transformations made using the specification of the present invention are included within the scope of the patent protection of the present invention.
Claims
1. A training sequence updating method for underwater acoustic communication channel estimation, characterized in that: The following steps are included: Channel coding: At the transmitting end, the bit stream to be transmitted is mapped into symbols and then divided into N groups on average. Each group is divided into M blocks on average. First, CRC coding is performed on each block of symbols. Then, error correction coding is performed on the entire group of symbols after processing. Then, the error correction code check symbols are evenly divided into N blocks and dispersed to each block to form new blocks. Finally, the information code symbols of each new block are interleaved with the CRC check code symbols and the error correction code check code symbols to complete the channel coding process. Signal frame design: take the first block of symbols in each group after channel coding and perform symbol mapping and signal modulation in sequence to form N blocks of coded signals for the first half frame, and perform signal modulation in sequence for the second block of symbols to form N blocks of coded signals for the second half frame, insert a training sequence signal between the first and second half frames to form the first frame of communication signal; take the coded signal generated by the third block of symbols in each group as the first half frame, and the coded signal generated by the fourth block of symbols as the second half frame, insert a training sequence signal between the first and second half frames to form the second frame of communication signal; and so on; Pilot sequence update: at the receiving end, channel estimation is performed on each training sequence signal of the M / 2 frame signal at the same time to obtain M / 2 channel transfer functions. A block of coded signal is taken before and after each training sequence signal, and channel equalization, demodulation, and deinterleaving are performed in sequence using the channel transfer functions of each frame. Error correction decoding is then performed on the M blocks of deinterleaved results. CRC check is then performed on each decoding result. The correct block information codewords and CRC check codewords are interleaved as a new training sequence to achieve training sequence update. Similarly, the new training sequence is used for channel estimation to obtain the channel transfer function, and a block of coded signal is taken forward and backward to repeat the above process until all code blocks are decoded.
2. The training sequence updating method for underwater acoustic communication channel estimation according to claim 1 is characterized in that: In the signal frame design, a total of M / 2 frame signals are finally formed, and each frame contains 2N block coding signals.
3. The training sequence updating method for underwater acoustic communication channel estimation according to claim 1, characterized in that: During the entire decoding process, if a CRC check error occurs, a block that is closest to the signal block to be decoded and has a correct check is selected as the training sequence.