Research On Blind Recognition Of Interleaver And Convolutional Encoders Based On Algebraic Theory

Blind identification and recognition of channel encoder, as the first procedure in obtaining information sink from the interleaved coded sequence, is regarded as an im-portant technology in digital communications. Either for cooperative or non-cooperative communications, blind identification technology is of great value in such fields as adap-tive modulation and coding (AMC), smart jamming, network attack, communication reconnaissance. Also, it can greatly promote the development of communication coun-termeasure system and AMC. Nevertheless, research on this technology is far from enough, thus carrying a thorough study of it is of great necessity and also significance.This research is conducted based on algebra theory and special coding structures. First, it points out the significance and also focus of the research, after making a litera-ture review of the technology. Second, an overview of algebraic theory and basic con-cepts about interleavers and convolutional encoders related to the blind identification technology is made, making relevant theoretical preparations for the follow-up study. Third, algorithms for cases of interleavers and convolutional encoders are proposed and testified one by one, in an effort to check each algorithm’s accuracy and practicality.This thesis has achieved the following innovations.Firstly, it creates blind identification of matrix interleaver and convolutional in-terleaver. The phase by phase method based on the binary data matrix transformation theory involving rank criterion and characteristics of linear code, by using matrix op-erations, is applied to estimate matrix interleaving parameters in noisy circumstance. Moreover, a novel joint detection method is created in a response to the scenario that no integrated code words exist in one period of convolutional interleaver.Secondly, a position-determined algorithm is proposed to blind identify complex helical interleaver. In relating to solve the problems that the current interleaver identifi-cation algorithm only applies to simple structure or fixed parameters, a new algorithm is established to obtain complex interleaver. Based on rank criterion, with the help of structure properties of helical scan interleaving as well as position-determined al-gorithm, blind identification of parameters is achieved in non-error circumstance. By combining position-determined algorithm and candidate parameters method, dimension of candidate range can be reduced and blind identification of helical interleaver in noisy circumstance realized. More importantly, the method with different structural charac-teristics, can be applied to the blind identification of different types of interleaver.Finally, a WHT-BM algorithm is created for the blind identification of rate k/n convolutional encoders. In responsive to problems that existing recognition algorithms has limited application to specific rate case, low fault tolerance, and high complexity, WHT-BM algorithm, with its fault-tolerant features, is applied to blind identification of any rate case. Based on the structure properties of convolutional encoders, constraint relation between code and parity check matrix is split in to n-k equations system of parity check vectors. Combining with Block Matrix method, WHT-BM utilizes all its coefficients and segments equation evolving coefficient matrix and parity check vector, which is recovered from the peak position in Walsh spectrum, and achieves the goals of reducing the computation dimension and substantially increasing the algorithm’s accuracy. After reconstitution, WHT-BM succeeds in the blind identification of rate k/n convolutional encoders in a noisy context.