Design LDPC-BICM Scheme Based On The Density Evolution

In recent years, as the fast development of wireless communication, people’s needs for highspeed and superior quality communication are increasing, and their demand for communicationsystem is getting higher and higher. The combination of the excellent performance of LDPC codeand high bandwidth-efficient of BICM system meets the requirement for high data rate and highreliability information transmission of modern communication.This article firstly makes a systematic study of the basic concept of LDPC code, theconstructing procedure of check matrix and encoding and decoding algorithms. Then it introducesBICM system model and MLC system model, and their principles, and compares these two models.It also analyses BICM channel capacity and gives sub-channel capacity simulation figure of256QAM under gray-mapped.Then it introduces the principle and algorithm of the density evolution of LDPC code, thealgorithm of the density evolution of Gaussian Approximation, also the initialization and how tojudge the results of the density evolution of BICM system of LDPC code. The algorithm of thedensity evolution of Gaussian Approximation only needs to consider the evolution ofone-dimensional average and it is much easier than the algorithm of the density evolution.And itsimulates the sub-channel initial LLR pdf of256QAM of gray-mapped under the principle of equalcapacity in matlab environment.At last the design plan of BICM of LDPC code which is based on the density evolution isproposed. Under the principle of equal capacity, the simulation results indicate that it is possible tomake the DE plan better than the current T2plan. It analyses and compares the probability densityfunction of the initial LLR messages respectively under the principle of equal capacity and actualstatistics. Then the probability density function of the initial LLR messages under actual statistics isapplied in the density evolution, it then compares the BICM system performance of these twodesign plans respectively under the principle of equal capacity and actual statistics. The simulationresults show that the calculated noise threshold based on the actual statistics is reduced by1dB thanthe actual Link Simulation, but for the BER performance of actual Link Simulation, the design planbased on the principle of equal capacity is much better.