Research On Coding Algorithms Of LDPC Codes

As for modern communication technology,especially in wireless communication and deep space communication field, channel coding holds an important status.As an important branch of channel coding, though LDPC is discovered early,until recently does its research have developed deeply.Compared with traditional coding methods,LDPC has significant advantages,for example,LDPC is suitable for parallel decoding,it has high real-time performance and the computational complexity of its encoding and decoding is low,etc.This article maily aims at this aspect:There exists short cycles in the random check matrix got accoding to traditional coding methods.Baesd on the relation between bipartite graph and check matrix,this article puts forward some improved algorithms.The mainly work in this article is as follows:1. Consult a great deal of literature and datum so as to understand the development and present situation of LDPC.2. Introduce the basic knowledge of LDPC.Firstly,introduce the encoding and decoding of binary LDPC.Then introduce the encoding and decoding of m-ary LDPC.What’s more,introduce the basic knowledge of graph theory,then futher explain the geometrical significance of cycles that exist in check matrix and their harm.3. Discuss the improved algorithms of traditional coding methods for random LDPC.1) After illustrating some traditional coding methods for random LDPC,take Gallager’s method for example,put forward a method to remove cycles(length4).Further more,put forward a method to limit cycles(length4) in check matrix.2) Put forward similar methods as for cycles(length6) on the basis of having removed cycles(length4).4.Finally, carry out simulations so as to test and verify validity and advancement of these improved algorithms.