Research On The Algorithm Of Convolutional Code Blind Recognition

In the digital communication system,the channel coding technology ensures the reliability of the transmitted information and overcomes the interference of noise in the channel.In a non-cooperative communication scenario,the interceptor cannot directly learn the channel coding scheme and parameters used in the communication process.Therefore,there is a channel coding identification technology that recognizes the coding parameters from the received bit stream without or only partial prior knowledge,and supports the next decoding operation to obtain reliable information.This technology plays an important role in the fields of communication confrontation and intelligent communication.Therefore,the research of this technology has important military significance and application value.For system convolutional codes,the existing blind identification algorithm is mainly for convolutional codes with common code rates of 1/n and(n-1)/n,and fault tolerance performance and computational complexity cannot be taken into account.The generation matrix of the non-systematic convolutional code has a one-to-many uncertainty relationship with the check matrix.At present,there are few studies on the blind identification of nonsystematic code coding parameters.In summary,the research on blind identification of nonsystematic convolutional code coding parameters at arbitrary bit rate has important practical significance and application value.This paper studies the blind identification problem of arbitrary code rate convolutional code coding parameters.Firstly,with regard to the identification of the code length,the value of the code length and its multiple is filtered by the rank criterion characteristic,the code length traversal range is reduced,and then the analysis matrix is built by traversing the filtered value,and a fast Walsh-Hadamard transform is performed,using Walsh spectrum.The spectrum peak information of the spectrum identifies the code length information.After that,the matrix identification model is established according to the code length.We put the code bit stream into the model in turn.The preprocessing is performed to filter the code sequence with strong correlation,and the statistical method is utilized to identify the length of information bits.The method of the article improves the fault-tolerant performance.Then,this paper transforms the check vector identification problem into the search problem of lowcode heavy code words in the matrix.The received sequence constructs the analysis matrix according to the code length,and uses the improved Canteaut-Chabaud algorithm.The iterative algorithm recognizes the check vector.Compared with the traditional algorithm,the check vector identification algorithm has good fault tolerance performance.After that,for the identification of the system code generation matrix,in this paper,based on the one-toone relationship between the check sequence and the generator matrix,the Gaussian transform is used to identify the register length and the generator matrix.For non-systematic code generation matrices identification,check vector equations were established using solutions of equations fast Walsh-Hadamard transform,quickly and accurately identify the complete set of polynomials to generate and register length of the generator matrix obtained by screening the correct conditions in this paper.Finally,the algorithm theory of the proposed algorithm is simulated and analyzed.The results show that the algorithm can complete the blind identification of arbitrary rate convolutional codes,and can achieve better recognition results and meet the requirements of blind recognition in non-cooperative scenarios.