The Design Of High Performance LDPC Codes And Its Application For Cognitive Radio Networks

ABSTRACT:Low Density Parity Cheek (LDPC) code have attracted considerable attention in coding communication because they can achieve near-Shannon limit performance, and LDPC code have been widely used in satellite digital video, deep space communication, and wireless communication. With the wireless mobile communication development, people continue to put forward new demands to channel coding. Recently an increasing number of communication standards are aimed at the LDPC codes, in order to adapt to the needs of future wireless communication system. To solve the problem of reliable data transmission under conditions of low signal-to-noise ratio, LDPC codes have been a standard channel coding in standards: digital video broadcasting standard GB20600,DVB-S2, broadband wireless access standard IEEE802.16e and deep space communication standard CCSDS. There are three requirements for the design of high performance LDPC codes:1. low coding complexity;2. fast encoding and decoding;3. good BER performances. Protograph LDPC codes are a class of structurized LDPC codes. Protograph LDPC codes have the advantage of high-speed decoding, low error floor and low iterative decoding threshold. Through the combination of theoretical analysis and computer simulation, this thesis studies the constructing of protograph LDPC codes, and focuses on the fast coding, coding complexity and BER performance and other aspects. And gives the theorem and corollary, and proves them. Some key technologies of cognitive radio networks based on high performance LDPC codes are also studied. The main works are as follow:1. The necessary and sufficient condition of protograph LDPC codes without four cycles and corresponding check algorithms based on index matrix are proposed. They can be used to analyze and optimize the design of protograph LDPC codes.2. The necessary and sufficient condition of protograph LDPC codes with nonsingular check matrix is proposed, and solves singularity problem of check matrix. The proof of circular generator matrices is given, and effectively solve singularity problem of check matrix, and ensure that the generation matrix having the cycle characteristics. It is provides a theoretical basis to construct high performance LDPC codes. Three construction methods of high performance LDPC codes with circular generator matrices and good BER performances are proposed.3. Fast iterative encoding algorithm based on check matrix is proposed. It can reduce the coding complexity. Two construction methods of high performance LDPC codes based on fast iterative encoding with linear complexity and good BER performances are proposed. Time-frequency code based on high performance LDPC codes is proposed.4. Some key technologies of cognitive radio networks based on high performance LDPC codes are also studied. Cognitive radio networks interference cancellation method based on high performance LDPC codes is proposed. It solves the problem of interference between the primary user and cognitive user. Three design schemes of cognitive radio networks based on high performance LDPC codes are proposed. A secure cognitive radio network using spreading LDPC codes is proposed.