Optimization Design And Application Of Fountain Codes

With the rapid development of multimedia broadcast and multicast business in the mobile Internet, the reliable transmission of data has become a hot research area. In addition, in order to maximize the limited spectrum resources, relay and cognitive radio technology has also attracted more and more attention. As an application layer forward error correction technique, fountain code can be effectively against co-channel interference, and is suitable for reliable broadcast and multicast transmission in cognitive relay network. Focus on fountain codes, this thesis has made some improvements in the optimization of encoding construction as well as in practical application promotion.This paper first reviews the channel coding theory and its development history, and introduces the evolution of the fountain code, its advantages/disadvantages and application scenarios. Then we introduce the fundamental theory, encoding and decoding algorithm of typical fountain codes such as Tornado codes, LT codes and Raptor codes. We also analyze the important parameters that affect the encoding performance, and successively analyze the advantages and disadvantages of these fountain codes.Secondly, according to the encoding/decoding fundamental of fountain codes, we propose a fountain code optimization construction method based on chaotic sequence. Simulation results show that, compared with the existing LT codes, our optimization construction method has a smaller bit error rate and a higher decoding success probability, effectively improve the coding efficiency.Finally, in order to improve the spectrum utilization, we apply the optimization design of fountain codes in cognitive relay system. By the simulation analysis, we determine the key parameters that make the secondary users get the best spectrum efficiency. The main contributions and innovation of our research lie in the following areas:In fountain codes performance improvements, we take advantage of the chaotic sequence characteristic, so that the parameters in the LT code structure matrix can be fixed in advance. We also change the selection mode of the adjacent nodes, and calculate the optimization parameter in the matrix according to the theory of the "And-or Tree", improving the decoding success probability even when code length is short. For fountain codes application, a cognitive relay network data transmission model using fountain codes as channel coding is proposed, and a secondary user spectrum efficiency formula is given. After model equivalent and simplify, we obtain the key parameters that affect the spectral efficiency, and maximize spectrum utilization.