Research of Wireless Fading Channel Modeling Based on Radial Basis Function Network

Radial basis function network (RBFN) is one of theneural networks used widely. Aimed at the complicated and nonlinear relationship between input and output property of wireless channel and the advantages of RBFN, a method for wireless channel modeling and simulation based on RBFN is presented in this paper. We construct the structure of RBFN that used for wireless fading channel modeling, and adopt the K-nearest neighbor algorithm and least square method to train the network. We discussed the fading channel model and analyzed the impact factor of little-scale fading channel modeling. With the ability of strong function approach and fast convergence of RBFN, the modeling method can implement the modeling and simulation of fading channel rapidly and effectively by learning the propagation characteristic information of wireless channel. The simulation result shows the feasibility and validity of modeling method.