Research On Blind Recognition Of Linear Block Code

The purpose of blind recognition of channel coding is to indentify the channel coding parameters when the intercepted bit stream is known with little or no prior knowledge.With the wide application of channel coding technology,blind recognition of channel coding parameters has more important application value in intelligent communication and information interception fields.This thesis focuses on the research of the binary linear block code,binary cyclic code in the hard decision case,binary cyclic code in the soft decision case,the main work is as follows:?1?Aiming at solving the problem that fault tolerance of blind recognition of linear block code parameters is poor,a blind recognition algorithm of linear block code parameters based on iterative column elimination method is proposed.First of all,a matrix is filled with intercepted bit stream,and then an iterative column elimination algorithm is applied to the intercepted matrix.The corresponding column in the transfer matrix of a dependent column in each window is regarded as candidate parity-check vector.In addition,the length of codeword and synchronization are estimated when the normalized dual space dimension reaches the maximum.Finally,through elementary row transformation of dual codewords,parity check matrix can be known.?2?Aiming at solving the problem that the blind recognition correct rate of cyclic code parameters is low and lots of intercepted data are needed,a blind recognition algorithm based on parity check matrix matching for cyclic code is proposed.First of all,the corresponding parity check matrix of all of the code length n and the factor of xⁿ-1 is regarded as candidate parity check matrix.Second,a matrix is filled with intercepted bit stream.It multiples candidate check matrix to determine whether exists parity check matrix at different code length and synchronization or not.Finally,according to the corresponding polynomial of the existence of check matrix,code length,synchronization and generate polynomial are estimated.?3?Because the code for the demodulation output contains more valid information in the case of soft decision than hard decision,and the study of blind recognition of cyclic code parameters is in the case of hard decision,a blind decision algorithm for cyclic code parameters using soft decision is proposed in this thesis.The method utilizes the M₂/M₄ estimator to estimate the amplitude and noise variance of the signal effectively.The parity check matrix is established by the property of the cyclic code so that the number of parity check vector is greatly reduced.The mean log likelihood ratio of the posterior probability of the corresponding syndrome is obtained by traversaling the parity check matrix at different codeword length and synchronization.The proposed method makes full use of effective information of the received symbols and the number of candidate parity check matrix is small,so the method can realize the full blind identification of the cyclic code parameters under the condition of low SNR,and has certain engineering application value.