High-reliability dynamic two-stage channel coding method for high-speed underwater acoustic communication

Abstract

The invention relates to a high-reliability dynamic two-stage channel coding method for high-speed underwater acoustic communication, which comprises the following steps of: (1) dynamically selecting a coding mode according to the relationship between the length of data to be transmitted and a preset threshold: if the total bit number D of the data to be transmitted meets the condition that D is less than or equal to B * R1 * Rh, adopting a Hamming code and LDPC cascade coding mode; otherwise, an LDPC coding mode is adopted, the LDPC coding process in the step (4) is directly executed, B is the bit number borne by a single OFDM block, and R1 and Rh are the code rates of the LDPC and the Hamming code respectively; (2) inputting data packets to form independent coding units; (3) Hamming code coding is carried out, and LDPC coding input is generated; (4) carrying out LDPC coding to generate final coded data; and (5) data transmission and decoding.

Description

Technical Field

[0001] The invention belongs to the technical field of underwater acoustic communication channel coding and relates to a high-reliability dynamic two-stage channel coding method. Background Art

[0002] Underwater acoustic channels have characteristics such as low signal-to-noise ratio, strong multipath effects, large transmission delays, and rapid time variations. These factors pose a huge challenge to achieving highly reliable data transmission in underwater acoustic communication systems. Orthogonal frequency division multiplexing (OFDM) technology can effectively combat multipath effects and frequency-selective fading through frequency domain diversity, providing a practical high-speed communication solution for application scenarios such as underwater sensor networks and long-range unmanned underwater vehicles. Channel coding plays a very important role in OFDM underwater acoustic communication systems. This technology can generate corresponding redundant information according to specific rules, and then use this redundant information to identify and correct transmission errors. Low-density parity-check code (LDPC) is one of the best-performing coding methods and can greatly improve the reliability of system transmission in complex underwater acoustic channel environments.

[0003] In practical engineering applications, LDPC coding fully demonstrates its performance advantages when processing longer data blocks. However, for shorter data blocks, its performance is limited due to the inherent characteristics of its coding principle. Furthermore, to achieve rate matching, LDPC coding often requires zero padding or additional redundancy when encoding shorter data blocks. This processing not only introduces additional redundant bit overhead, resulting in unnecessary waste of system resources, but also reduces the utilization of precious bandwidth resources in the underwater acoustic channel. Summary of the Invention

[0004] In response to the above problems, the present invention proposes a highly reliable dynamic two-stage channel coding method for high-speed underwater acoustic communications. This method adopts a dynamic coding selection mechanism based on data length, and through the cascade architecture of Hamming code and LDPC, it achieves efficient utilization of communication resources and a significant improvement in transmission reliability. For short data block transmission scenarios, the cascade mode of Hamming code and LDPC coding is adopted to avoid the redundant overhead problem caused by direct zero padding in traditional schemes; for long data block transmission scenarios, LDPC coding is directly implemented to give full play to its coding advantages in large data volumes. This dynamic coding strategy based on data length adaptation not only greatly reduces the consumption of system redundant resources, but also significantly improves the transmission reliability of high-speed underwater acoustic communications.

[0005] A highly reliable dynamic two-stage channel coding method for high-speed underwater acoustic communication comprises the following steps:

[0006] (1) Dynamically select the encoding mode according to the relationship between the length of the data to be transmitted and the preset threshold: Assume that the data to be transmitted is d = [d1, d2, ..., d D ], if the total number of bits D of data to be transmitted satisfies D≤B×R l ×R h When , the Hamming code and LDPC concatenated coding mode is used; otherwise, the LDPC coding mode is used and the LDPC coding process of step (4) is directly executed, where B is the number of bits carried by a single OFDM block, R l 、R h are the code rates of LDPC and Hamming code respectively;

[0007] (2) Input data grouping to form independent coding units: Group the data to be transmitted into k-bit groups. If the total number of bits D cannot be divided by k, calculate the number of required padding bits P1 = k - (D mod k) and fill them until each group of data meets the k-bit requirement, forming m independent coding units.

[0008] (3) Hamming code encoding, generating LDPC code input: using the preset generator matrix G h Perform Hamming code encoding on each group of k-bit data and output n-bit codeword c i , and finally merged into the complete encoded data C h =[c1 c2…c m ], as the input of LDPC coding;

[0009] (4) LDPC encoding to generate the final encoded data: First, dynamically adapt the input data source based on the encoding mode. If Hamming code encoding is not performed, the original data stream is directly used. Otherwise, the Hamming code encoding result C is inherited. h ; Then the data is padded to the LDPC encoder input length threshold through the zero padding mechanism, and finally the LDPC generator matrix G is used l Complete the encoding and output the encoded codeword C l ;

[0010] (6) Data transmission and decoding.

[0011] Furthermore, the method of step (4) is as follows: for the transmitted data u, let its bit number be K. If Hamming code encoding is not performed, then u = d, K = D; otherwise, u = C h , K=m×n; according to the formula P2=B×R l -K calculation code rate is R l The number of bits required to fill the LDPC code, and add P2 zero bits to the end of the data to be sent u; after the filling is completed, let the data to be sent after filling be u p , whose length is B×R l, that is, the length of the padded data meets the number of input bits required for LDPC coding; then, the preset generator matrix G is used l To u p Perform LDPC encoding; the output data after LDPC encoding is calculated to be C l =u p G l ; The data after LDPC encoding is the final encoded data, ready for data transmission.

[0012] Furthermore, the method of step (5) is as follows: the encoded data is modulated by OFDM and transmitted through the underwater acoustic channel. The receiving end recovers the frequency domain symbols through synchronization, channel equalization and fast Fourier transform (FFT) demodulation, and then dynamically selects the decoding mode according to the data frame header identifier, performs LDPC iterative decoding or cascaded Hamming code post-verification, and finally removes the padding redundancy to restore the original data.

[0013] The present invention provides a highly reliable dynamic two-stage channel coding method for high-speed underwater acoustic communication, aiming to improve the data transmission efficiency and reliability in OFDM underwater acoustic communication systems, especially to avoid bandwidth waste caused by zero padding when the data length is short. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 This is a block diagram of a high-reliability OFDM underwater acoustic communication system based on dynamic two-stage channel coding. DETAILED DESCRIPTION

[0015] The present invention will be described below with reference to the accompanying drawings.

[0016] The high-reliability dynamic two-stage channel coding method proposed in this invention combines Hamming code and LDPC code. Its main purpose is to improve the reliability of data transmission in OFDM underwater acoustic communication system and effectively solve the padding problem when the amount of data is small. Assume that the total number of bits sent in each OFDM block is B and the LDPC code rate is R l , and uses (n,k) Hamming codes. Hamming codes combine information bits with redundant bits to form reliable error-correcting codes, while LDPC codes provide powerful error correction capabilities, further improving the reliability of the communication system.

[0017] The specific operation process of the present invention is as follows:

[0018] (1) Coding mode selection: First, assume that the data to be transmitted is d = [d1, d2, ..., d D ], where D is the total number of bits of the input data to be encoded. For each OFDM block, we first need to determine whether the length of the data to be sent can be encoded by Hamming code. The specific judgment method is: calculate whether the number of bits to be sent D satisfies D≤B×Rl ×R h , where R h = k / n is the encoding rate of the Hamming code. If this condition is met, then Hamming code encoding can be performed. Otherwise, go directly to step (4) to perform LDPC encoding.

[0019] (2) Input data grouping: If the total number of bits of the input data is not enough to form a complete group, zero padding is required until the total number of bits meets the grouping requirements. Specifically, based on the number of bits to be sent D, it is determined whether the requirement of k bits per group is met. If the number of bits D of the input data cannot be divided by k, the last group of data is zero-padded until the number of bits per group of data is k. The required number of supplementary bits can be calculated as P1 = k-(D mod k). After zero padding, the input data will be evenly divided into m groups, each with k bits. The data after zero padding can be expressed as: {d1, d2, ..., d k},{d k+1 ,d k+2 ,...,d 2k},…,{d (m-1)k+1 ,d (m-1)k+2 ,...,d mk}, each group of k bits will be used as an independent coding unit.

[0020] (3) Hamming code encoding: The generator matrix of (n, k) Hamming code can be expressed as G h =[I k P], where I k is a k×k order unit matrix, P is a k×r order matrix, and r=nk is the number of supervision bits. Then for each set of input data d i =[d (i-1)k+1 ,d (i-1)k+2 ,...,d ik ], use Hamming code to encode it and get the corresponding output codeword c i =d i G h , where c i =[c i1 ,c i2 ,...,c in ] is the encoding result of length n. By performing Hamming code encoding on all groups, the complete encoded data C is obtained. h =[c1 c2…c m ], which is used as the input data for the subsequent LDPC encoding.

[0021] (4) LDPC coding: For the data u to be sent, let its number of bits be K. If Hamming code encoding is not performed, then u = d, K = D; otherwise, u = C h, K=m×n. Next, according to the formula P2=B×R l -K calculation code rate is R l The number of bits required to fill the LDPC code and append P2 zero bits to the end of the input data. After filling, let the filled data be u p , whose length is B×R l , that is, the length of the padded data meets the number of input bits required by LDPC coding. Next, use the preset generator matrix G l To u p Perform LDPC encoding. The output data after LDPC encoding is calculated as C l =u p G l The LDPC-encoded data is the final encoding result and is ready for data transmission.

[0022] (5) Data transmission and decoding: The LDPC-encoded data is distributed to multiple orthogonal subcarriers through OFDM modulation, forming a time domain signal and then transmitted through the underwater acoustic channel. At the receiving end, synchronization and channel equalization are first performed on the received signal, and the coded symbols on each subcarrier are restored through FFT demodulation. The decoding mode is then dynamically selected based on the data frame header identifier: If it is a Hamming-LDPC cascade coded frame, a belief propagation (BP) decoder is constructed based on the Tanner graph, and LDPC decoding is implemented by iteratively updating the probability information of the variable nodes and the check nodes until the maximum number of iterations t is reached. max Or satisfy the parity check equation If the identifier indicates a pure LDPC coded frame, the LDPC decoding sequence is directly output. For the data frame of the concatenated Hamming code, the Hamming code syndrome detection is further performed. If s≠0, locate the error bit position by looking up the table and perform error correction. Finally, remove the zero padding bits added during the encoding process and restore the original data sequence d=[d1,d2,...,d D ] to achieve highly reliable data transmission.

Claims

1. A highly reliable dynamic two-stage channel coding method for high-speed underwater acoustic communication, comprising the following steps: (1) Dynamically select the encoding mode according to the relationship between the length of the data to be transmitted and the preset threshold: Assume that the data to be transmitted is d = [d1, d2, ..., d D ], if the total number of bits D of data to be transmitted satisfies D≤B×R l ×R h When , the Hamming code and LDPC cascade coding mode is adopted; otherwise, the LDPC coding mode is adopted and the LDPC coding process of step (4) is directly executed, where B is the number of bits carried by a single OFDM block, R l 、R h are the code rates of LDPC and Hamming code respectively; (2) Input data grouping to form independent coding units: Group the data to be transmitted into k-bit groups. If the total number of bits D cannot be divided by k, calculate the number of required padding bits P1 = k - (D mod k) and fill them until each group of data meets the k-bit requirement, forming m independent coding units. (3) Hamming code encoding, generating LDPC code input: using the preset generator matrix G h Perform Hamming code encoding on each group of k-bit data and output n-bit codeword c i , and finally merged into the complete encoded data C h =[c1 c2…c m ], as the input of LDPC coding; (4) LDPC encoding to generate the final encoded data: First, dynamically adapt the input data source based on the encoding mode. If Hamming code encoding is not performed, the original data stream is directly used. Otherwise, the Hamming code encoding result C is inherited. h ; Then the data is padded to the LDPC encoder input length threshold through the zero padding mechanism, and finally the LDPC generator matrix G is used l Complete the encoding and output the encoded codeword C l ; (5) Data transmission and decoding.

2. The high-reliability dynamic two-stage channel coding method according to claim 1, characterized in that: The method of step (4) is as follows: for the data u to be sent, let its bit number be K. If Hamming code is not performed, then u = d, K = D; otherwise, u = C h , K=m×n; according to the formula P2=B×R l -K calculation code rate is R l The number of bits required to fill the LDPC code, and add P2 zero bits to the end of the data to be sent u; after the filling is completed, let the data to be sent after filling be u p , whose length is B×R l , that is, the length of the padded data meets the number of input bits required for LDPC coding; then, the preset generator matrix G is used l To u p Perform LDPC encoding; The output data after LDPC coding is calculated as C l =u p G l ; The data after LDPC encoding is the final encoded data, ready for data transmission.

3. The high-reliability dynamic two-stage channel coding method according to claim 1, characterized in that: The method of step (5) is as follows: the encoded data is modulated by OFDM and transmitted through the underwater acoustic channel. The receiving end recovers the frequency domain symbols through synchronization, channel equalization and fast Fourier transform (FFT) demodulation, and then dynamically selects the decoding mode according to the data frame header identifier, performs LDPC iterative decoding or cascaded Hamming code post-verification, and finally removes the padding redundancy to restore the original data.

Images