Analysis And Optimization Of Iterative Decoding Performance In Powerline And Wireless Cooperative Communication

Powerline communication and wireless communication have broad applications prospects in smart grids and smart homes.The integration and cooperation of the two can significantly improve the utilization of communication system resources and the level of network intelligence,which can provide users with diverse communication services.Aiming at the mixed fading and impulse noise environment of power line and wireless dual-media cooperative communication systems,this paper studies an algorithm based on Gaussian mixture(GM)model which can analysis low-density parity check(LDPC)codes through multidimensional external information transfer(EXIT),solves the problem that the traditional EXIT algorithm is not applicable to non-ideal channel.In order to solve the problem that the traditional EXIT chart is not applicable to the non-ideal channel conditions,this paper proposes a multi-dimensional EXIT theory based on GM model for the channel with mixed fading and impulse noise.First,the dual-media cooperative communication system is modeled,and the maximum ratio combining(MRC)is used to process signals containing the same information from different channels at the receiving terminal.Then,the probability distribution of the log-likelihood ratio(LLR)of the received signal at the receiving terminal is decomposed by Gaussian mixture model,and the key parameters of the Gaussian component are estimated by expectation maximum(EM)algorithm.Based on Gaussian mixture model,the traditional EXIT mutual information calculation formula is improved to solve the problem that the variances of the Gaussian components are not twice of the means.Finally,simulation analysis verifies the effectiveness and reliability of the multi-dimensional EXIT algorithm in powerline and wireless cooperative communication systems.The traditional construction methods of LDPC codes commonly used in Gaussian channels may not be suitable for powerline and wireless dual-media cooperative communication systems.A joint optimization algorithm of LDPC degree distribution considering both noise threshold and decoding error rate is proposed in order to solve the problem of LDPC code word construction in the channel with mixed fading and impulsive noise.First,the stability and convergence conditions in the multidimensional EXIT algorithm are derived and proven.Next,the stability and convergence conditions are used as constraints,and Differential optimization of LDPC codes to maximize the area between the curves and the lowest bit error rate when converging.Finally,a series of comparative simulation verification is performed on the optimized degree distribution.The simulation results show that the theoretical bit error rate and convergence threshold of the LDPC codes obtained by the above optimization algorithm are better than the common regular and irregular codes.