Modeling And Simulation Performance Evaluation On DMC-based Geometric Random Clustering Channel

As wireless communication technology and people’s lives more and more integral,intelligent terminals can obtain multidimensional environmental information. How totake advantage of access to environmental information to improve spectral efficiencyand data transfer rates, so as to enhance the quality of communication is a problemthat must be addressed today. Channel modeling is the most simple and cost-effectiveway for algorithm validation.This dissertation introduces MIMO channel modeling methods, by comparingexisting geometric random clustering channel model, on the basis of previous model,increasing the scattering multipath component (DMC), which has a correlation withdiscrete components. The model receive data extracted the frequency domain, timedomain and space domain, angle domain, Doppler spread and noise characteristics ofmultidimensional environmental information by using terminal. According tomultidimensional and multi-scale extraction of dissemination in the dynamic randomwaves of wireless environment establish wireless channel modeling environment. Byincreasing the DMC component consideration to improve system capacity, increasethe accuracy of the system. The advantage of establishing cluster model based onDMC geometric random is that not only environmental characteristics can be as inputparameters by forward simulation model, but also parameter estimationalgorithm(SAGE) can be to get from the measurement data and then accept the modelparameters as verified channel inversion.In this dissertation, setting environmental parameters, establishing forward model andinferring mathematical expressions are based on DMC geometric random clusteringmodel, and using SAGE algorithm extracted parameter through signal the channelgenerated, taking a comparative study analysis under different scenarios, testingmodel’s scene ranges. With indicator of environmental testing system parameters,giving out simulation result for setting of the geometric random clusters and PDFdistribution of each parameter, verifying the effectiveness of the established model.