The Study And Implementation On DSP For Channel Coding Of The Underwater Acoustic Communication

All activities that people take part in to develop and utilize marine resource, are all closely linked with underwater acoustic communication technology. Through constant development, underwater acoustic communication system has formed a complete system.However, the underwater acoustic channel has characteristics such as narrow frequency band,strong multipath effect, so the bit error rate of un-encoded underwater communication system is higher. It is usually difficult to meet the actual demand. It is of great significance to achieve high quality of underwater acoustic communication that the study of channel coding technology. By adding check squence into the information sequence, channel coding technology allows the receiver to correct error in receive sequence. It's significative to reduce the error caused by channel interference, to reduce the bit error rate of underwater acoustic communication. This article is mainly on the Matlab simulation research and the DSP implementation about Reed-Solomon (RS) code, convolution code and Turbo code of channel coding.The first major part is about the Matlab simulation of the channel coding. Firstly, the simulation of RS code is introduced. Mainly by changing two parameters, the code length and the error correction ability, the factors affacting the bit error rate are discussed. Then the simulation of convolution code is represented, majoring on the influencing factors of convolution code bit error rate, such as the bit rate, generator matrix, constraints length,tracing back length, ruling way and the quantitative value of Q in soft decision. On the simulation of Turbo code, not only four parameters are analyzed, namely the bit rate,component code, number of iterations and weaving length length. But also three decoding algorithms of the Turbo code, Log_MAP, Max_Log_MAP and SOVA, are compared at the same time, in order to understand the influence of different decoding algorithm on performance. Finally three kinds of channel coding methods are compared under white gaussian noise channel. By adding impulse response of the pool channel for the calculation,performance characteristics of different kinds of channel coding under the underwater acoustic channel are analysised.Another major part of the paper is the realization of the three kinds of channel coding on the DSP. The hardware platform of the software implementation is the DSP chip TMS320C6455. Firstly, detailed introduction of software implementation process of the coding and decoding is carried on, about the RS code, convolution code and Turbo code. It's mainly described through the form of program flow chart, while the allocation of storage space of variables is given during the decoding process. Then by adding some errors to the encoded data, with the comparison of the data before encoding and the data after decoding,it's verified that three kinds of coding are implemented on the DSP successfully in the paper.Finally, by consuming time of coding and decoding of three kinds of channel coding, the algorithm complexity is compared in the paper.Through the results of the simulation and the implementation on the DSP,we can conclude that, in terms of error correction performance, the performance of the Turbo code is much better than RS code and convolution code. But in terms of implementation complexity,the complexity of Turbo code is higher than the other two coding methods. In underwater acoustic communication system, Turbo code is a solid choice in order to improve the reliability of the information transmission, to satisfy the demand of the high speed data communication and the efficient resource utilization.