Research Of Blind Identification Of Channel Coding Technology Based On Sparse Decomposition In Finite Fields

With the development of digital communications technology, the channel coding technologycan increase the stability and reliability of data transmission, has been widely used in the field ofwireless communication, but also increased complexities of analyzing and processing the receivedinformation. If the receiver is able to identify channel coding automatically, it will greatly enhancethe ability of adaptive communication and improve the frequency spectrum utilization and datatransmission rate. The channel recognition technology is widely used in noncooperative signalintercepting, cooperation communication, intelligent mobile communication, multicastcommunication and so on. So it’s not only of great importance in theoretic but also of greatsignificance in practical applicating to do in-depth search in this field.Sparse decomposition did atomic decomposition by searching matching base functions in theatomic dictionary. It reflected the signal information on the sparse components. So it has many verygood characteristics and has been widely applied in signal classification, medical imaging, superresolution radar, ultra broadband channel estimation and so on. The applications of blindidentification of channel coding are special, so there are few published relevant literatures. In thispaper, drawing on previous researching of blind identification, we take the sparse decompositiontheory into blind identification of channel coding, and propose a new method of blind identificationbased on sparse decomposition.This paper mainly includes several parts: First of all, we researched the blind identification oflinear block code. We recognized the encoding length and starting position of the linear block codesby doing linear transformation of the matrix and solutioning its rank. Then we built the analysismatrix and simplified it and identified generator matrix and generator polynomial. Then weresearched the blind identification of cyclic code and recognized the encoding length and startingposition by the same method. The generation polynomial can be identified by calculating thegreatest common divisor of cyclic codes. Relative to convolution code, we constructedover-complete dictionary according to the prior check matrix, and judged whether the codes hadmatched over-complete dictionary and recognized the convolution code. We used examples toverify its feasibility and validity and analyzed the performance of the algorithm. Finally, sparsedecomposition was used to blind identification of channel coding. By studying constraint relationsof the code sequence between the encode elements and we gave a unified expression for theunknown constraint relations. It was converted to how to solve the constraint relationship. In fact, itwas a problem about sparse decomposition in finite field. Encoding methods and parameters can be recognized by solving the convex optimization problem. We also used examples to verify itsfeasibility and validity.