Blind Recognition And Parameter Estimation Of Cyclic Codes

The purpose of channel coding recognition is to estimate the coding parameters ofthe intercepted code sequence with just a few or even without prior information.Channel coding recognition is a typical signal reverse acquisition problem. With thecontinuous development of the digital communication technology, people have madehigh demands on the adaptability of the communication system. Channel codingrecognition technology is the foundation of the realization of intelligent communication.Besides that, it is the premises, which communication surveillance can go deep into theinformation layer.Cyclic code is a kind of commonly used channel code. The research centered onthe blind estimation of parameters of cyclic codes. The main contents are as follows:(1) Some basic concepts and important properties in abstract algebra are brieflyintroduced to lay the foundation for the following discussing.(2) After the analysis of the mathematical properties of cyclic code, an universalalgorithm for the blind estimation of cyclic code is presented. First, the code length isobtained by using rank rule method. Then by determining the factors step by step, thegenerator polynomial is estimated. The simulation results show that the proposedmethod has a identification rate over80%even when the BER is2%.(3) A blind estimation algorithm for primitive BCH code is presented. Making fulluse of the efficient decoding algorithm and strict algebraic structure of BCH code, theproposed method can determine some factors of the generator polynomial whileestimating the block length, which greatly improve the recognition efficiency. Thesimulation results show that the performance of algorithm for primitive BCH code isbetter than the universal method.(4) A blind estimation algorithm for RS code is presented. In binary domain, RScodes are quasi-cyclic codes. Making use of the structural features of the equivalentbinary quasi-cyclic code of RS code, the proposed method combines the estimation ofblock length and primitive polynomial. The computational complexity analysis shows that the proposed method has lower computational complexity than the traditionalmethod which is based on GFFT.