A Joint Design Of LDPC And Network Coding Basing On Information Compensation Of Information-correction

Nowadays, with the rapid development of wireless Internet, a greater amount of data and faster data transfer rate are very important in wireless communication system.However, the relay node is required to complete the data transfer due to fading characteristics of wireless systems.In the year of 2000, R.Ahlswede first proposed the concept of network coding,which means the relay node can process the data in some way before forward it to the destination node. The network coding technology using in wireless cooperative relay system can enhance network reliability, reduce energy consumption, improve network throughput and other advantages. At the same time, recent studies show that LDPC code which can approach the Shannon limit has also been more widely used in communication systems.Therefore the joint design of network coding and LDPC code can be an effective way against fading wireless channel.Based on the analysis of two simple joint design methods, this paper studies the joint design method of 16 QAM network coding and LDPC code.This method improvs the traditional mapping mechanism and improves the information transfer rate because of the high-order modulation.Simulation analysis shows that its performance is improved comparing to the method not using the physical layer network coding.Then we studied the joint design method basing on soft information.In this method.The relay node uses soft decision in order to avoid losing important information which is benefit to the decision of the destination node.But this method increased the complexity of the processing of the relay node.So we propose a new joint design method of network codes and the LDPC codes basing on information-correction compensation.In this method,the relay node decodes a hard decision and the destination node has an error compensation according to the channel characteristics.Simulation analysis shows that this system performance of the new joint design method is much better and also has a good performance between the system complexity and the system performance.