Type Identification Of Error Correcting Codes And Parameter Estimation Of Shortened RS Codes

In digital communication system,various interferences and noises in the channel lead to errors in the transmission process.In order to ensure the reliability of the transmission process and reduce the impact of error codes on the correct information,channel coding techniques are often used.The traditional encoding and decoding techniques are based on cooperative communication.The coding types and coding parameters are known,and the receiver can recover the information sequence and complete the information transmission by adopting an appropriate decoding method based on the prior information.However,in the case of non-cooperative communication,it is an important measure to obtain non-cooperative information in communication detection to identify the encoding type and estimate the encoding parameters under the condition that no prior information or only a small amount of prior information is known to the receiver.This thesis mainly focuses on the type recognition of coding techniques used for error correction and parameter estimation of shortened RS codes?First,this thesis gives the background and significance of the research topic,explain the research status of related technologies,summarizes the results of available research,and puts forward the research content and research framework of this thesis.Second,in order to focus on the research related to the type recognition of coding techniques,a set of error-correcting coding styles based on closed sets is selected.Different coding types include: Hamming codes,convolutional codes,primitive BCH codes,shortened BCH codes and RS codes,and the principles of these coding techniques are briefly explained.Then,the characteristics of code words for different coding types are analyzed to pave the way for further research.Third,for the type recognition technology of the error correcting codes based on closed set,the overall recognition framework is given: firstly,the code length recognition algorithm based on Gauss-Jordan elimination through pivoting(GJETP)method is applied to estimate the code word length,and then the types of the code are roughly classified by the estimated code word length.Based on the difference between convolution codes and linear block codes,a recognition method based on the correlation coefficient of runs statistical vector is proposed to distinguish the code words of the two coding types.In view of the difference between the finite field characteristics of shortened BCH codes and RS codes,the recognition algorithm based on the normalized spectrum of Fourier transform in Galois field is applied to distinguish the code words of the two coding types.Finally,to address the shortcomings,that the existing algorithm to finish parameter estimation of non-shortened RS codes cannot be directly applied to shortened RS codes,a new code word length estimation algorithm based on Euclidean method is proposed.This algorithm not only reduces the amount of data needed for recognition but at the same time achieve better anti-error performance.Then,the applicability of two common primitive polynomial recognition algorithms to shortened RS code words are compared and the recognition algorithm based on Galois Field Fourier Transform(GFFT)is selected as the primitive polynomial recognition algorithm in the overall recognition scheme.Finally,an overall scheme of shortened RS code recognition is presented,and verified through simulation results,which lays a foundation for further research.