Research Of Recognition And Analysis Technology Of LDPC Code

Channel Coding Recognition Analysis(CCRA)which is widely applied in intelligent communications and information countemmeasure,refers to the analysis coding type and coding parameters solution of the received data coding sequence under the condition of semi-blind or full-blind.Low-density parity-check(LDPC)codes are wideiy empoloyed in civil and military fields,such as satellite communications,deep space communications,microwave communications and so on,because of their simple decoding,error-correcting performance close to Shannon and conduciveness for high-speed information transmission.However,since little is known abroad about the key technologies and latest achievements of LDPC coding and recognition analysis and the research on LDPC coding and recognition technology has just started in China,with many problems to be solved,especially in the practical aspects,the research of LDPC coding recognition and analysis technology is an urgent requirement.In this paper,LDPC coding recognition analysis technology under non-cooperative conditions(semi-blind or completely blind coding parameters)is investigated.Firstly,the research background and significance of LDPC coding and recognition are introduced,and the current research status and existing problems are examined.The LDPC coding theory and recognition problems are analyzed and summarized.Then,based on the traditional channel coding recognition methods,three algorithms are proposed and analyed experimentally from presepective of the recognition of LDPC code length and bit rate,of check vector and the reconstruction of check matrix.The main research work is summarized as follows:(1)In the process of semi-blind recognition of some known prior information,an LDPC coding recognition algorithm based on the maximum mean square ratio(MSR)is proposed to solve the problem of low distinguishability of the existing LDPC coding recognition algorithm by constructing the logarithmic likelihood ratio of the test relation and analyzing its probability distribution characteristics with different coding parameters,a maximum mean square ratio(MMSR)decision maker is designed.The experimental results show that the algorithm can effectively improve the performance of the decision maker and solve the recognition problem of high bit rate LDPC codes in low SNR environment.(2)A new algorithm of LDPC code verification vector recognition based on sparse constraints is put forward in the process of full blind recognition without prior information,aiming at the high search complexity and low recognition rate of existing LDPC code verification vector recognition algorithms.By utilizing the sparse constraint property of check vectors of LDPC codes,the pre-search space of check vectors,which reduces the complexity,is constructed.Moreover,several groups of independent data are used to verify the iteration of the check vector,which improves the recognition rate of the algorithm.Experimental results indicate that the vector search space of the algorithm is significantly reduced and that the recognition rate of the algorithm is increased by 15% approximately,the decoding gain of the algorithm by 2.3 dB.(3)A blind recognition algorithm of the check matrix is proposed to solve the difficulty of reconstructing the check matrix under the condition of error code.First,the intercepted data are segmented and the error matrix is consttucted.The dual vectors are obtained by column elimination of the error matrix.Then,the valid code check vectors in the dual vectors are selected by combining the criteria of the check vectors.The error-free code groups in the intercepted data are identified and eliminated,and the ratio of the error-free code groups is improved continually.Finally,the thorny problem under the original error condition is reduced to the reconstruction problem of linear constraint relation under the error-free condition.The experimental results indicate that the algorithm can reconstruct the code check matrix to a certain extent when the bit error rate is fixed.