Blind Recognition And Parameter Estimation Of Nonsystematic Codes

Channel coding is an important part of digital communication system,while nonsystematic coding is a common form of channel coding.In many coding methods,convolutional codes have excellent performance and wide application.Therefore,this paper mainly focuses on the blind recognition of nonsystematic convolutional codes.In this paper,a recognition algorithm for nonsystematic convolutional codes at various bit rates is established.The specific contents are as follows:First,in the case of known bit rate,several algorithms are summarized for recognizing the polynomial generated by 1/2 convolutional codes: data matrix IP transform method,Grobner-based algorithm and Walsh-Hadamard transform method.The fault tolerance of various methods is analyzed,and an effective method of generating polynomial estimation is obtained.Then,the case of the bit rate extended to 1/n is considered.All parameters of convolutional codes needed to identify and estimate in this case.Bit rate estimation is divided into two parts: code length and information group length.In this paper,the Key Module Equation is generalized and the related algorithm for solving the closing die is proposed,which has a certain error tolerance.Several recognition algorithms are given for generating polynomial matrix and estimating the starting point of codewords,and the overall recognition process of convolutional codes is synthesized as well.Finally,the problem of identifying k/n rate punctured convolutional codes is solved.A new algorithm of recognizing arbitrary k/n convolutional codes is given by solving equations with equivalent system coding.The algorithm in this paper is suitable for nonsystematic convolutional codes with various bit rates,and can estimate all kinds of parameters completely.All kinds of algorithms can maintain good performance at a certain bit error rate,and can be applied to practical projects.