信道建模与仿真.ppt

Modeling and Simulating the Wireless Channels Prof. Li Pingan Wuhan University of Tech. Flat fading channel model Two classical simulation model early R.H. Clarke, Bell Syst. Tech. J., 1968 W.C. Jakes, Microwave Mobile Comm., 1974 Conventional flat fading model Jakes model Clarke Gaussian noise filter model M. Patzold fading model in cell mobile communication system Jakes model The sum of a set of sinusoids with uniformly distribution on phase Jakes model Improved Jakes model Improved Jakes model Gaussian noise filter The sum of two quadrature independent Gaussian random variables Suzuki process Suzuki process is a product process of a Rayleigh process and log-normal process PDF and auto-correlation for Jakes power spectral density function The autocorrelation function The mean power of Method of equal distances: Rayleigh-distribution PDF and auto-correlation for Method of equal distances: Log-normal distribution Suzuki Process Simulator Simulation Approaches for Suzuki Process 等距离法 Method of Equal Distances (MED) 等面积法 Method of Equal Area (MEA) 蒙特卡罗方法 Monte Carlo Method (MCM) 均方误差法 Mean Square Error Method (MSEM) Simulation Results by using MED 瑞利过程仿真 理论值和仿真值比较 理论值和仿真值比较 理论值和仿真值比较（相关函数） TDL (Tap-Delay-Line) Model X X X + X X X X + X (N=4M+2) where denotes the maximal Doppler frequency X X X + X X X X + X (N=4M+2) where denotes the maximal Doppler frequency, are independent random variables uniformly distributed over for all n. Independent complex Gaussian white noise source IFFT IFFT Complex Gaussian Gaussian Rayleigh Log-normal ? ? Power spectral density function The autocorrelation function 3dB cutoff F To limit the F-region X X X + X X X + X X X + X + s m X Rayleigh Process Log-normal Process Suzuki Process 对数正态过程仿真器 It can be seen that the simulation model coincides very closely to the theoretical model 理论模型: It can be seen that