Research On Interleaved Identification Technology Based On Statistical Characteristics

In non-cooperative communication,it is very important for the reconnaissance party to obtain the intelligence information of the reconnaissance party.However,the using of interleaving technology has brought great difficulties to the identification and analysis of the reconnaissance party.Interleaving identification is a key difficulty in the analysis of channel coding parameters.The study of interleaving identification technology has important practical value and demand significance in the field of non-cooperative communication.Therefore,this thesis focuses on the problem of random interleaving identification of convolutional codes.Fisrt,based on the analysis of the statistical characteristics of the convolutional code,this thesis completes the estimation of the interleaving depth and the weight of the check code of the convolutional code.And then realize the identification of the random interleaving permutation relationship between the encoder's register state of return to zero and non-return to zero in order.Specifically summarized as follows:1.Using the conclusions obtained from the experimental analysis of the statistical characteristics of the correlation columns of the convolutional code,the estimation of the interleaving depth and the weight of the check equation is completed.The validity of the estimation is proved by simulation.At 3‰bit error rate,the recognition rate of the interleaving depth reaches 100%,and the weight recognition rate of the check equation exceeds 90%.And the recognition performance can be further improved by increasing the amount of data.2.To solve the problem of non-return-to-zero random interleaving between frames,first research and implement the matrix Gaussian transform method,derive the algorithm complexity of the method and propose an improvement scheme to reduce the complexity.Simulation shows that at10^-4 bit error rate,the algorithm has improved the recognition rate is increased from 21%to 90%,and the upper limit of the adaptive interleaving depth is 500.Then research a method based on graph isomorphism.This method uses low weight codewords to search the set of check equations,and then classifies them according to the neighborhood profile.Then,the graph representation of check equations is introduced to establish a random interleaving check before and after.The relationship between the equations,the isomorphic relationship of the search graph,and the mapping relationship between the labels on the edges are used to effectively identify the permutation relationship,and the method is verified by simulation.3.Aiming at the problem of non-return-to-zero random interleaving between frames,this thesis proposes an identification method based on equivalent system recursive convolutional codes and Viterbi decoding.Based on the equality of the beginning code groups of each frame,the initial two positions are determined.Non-systematic codes are converted into equivalent systematic recursive convolutional codes.Through similarity comparison and the use of divide-and-conquer ideas,iteratively searches for subsequent code groups.During iterative search,Viterbi decoding is used to correct errors,which effectively prevents the accumulation of bit errors.Simulation experiments show that the method has good fault tolerance and does not decrease with increasing interleaving depth,and can effectively achieve random interleaving identification with an interleaving depth of 1000.At 2.5%bit error rate,the recognition rate can reach 92%using only 50 frames of data,and by increasing the number of data frames,the recognition rate can be further improved.