基于在线训练RBF网络的OFDM信道跟踪  

OFDM channel tracking based on RBF online-training

作　　者：刘晶晨[1] 臧小刚[1] 凌小峰[1] 宫新保[1] 

机构地区：[1]上海交通大学电子工程系,上海200240

出　　处：《信息技术》2009年第4期1-4,共4页Information Technology

基　　金：上海市国际科技合作基金项目(06SN07112)

摘　　要：提出一种结合了正则化技术及时间加权技术的递归最小二乘算法,用于径向基函数神经网络的在线训练及动态调整,从而解决OFDM系统中信道的动态跟踪问题。该算法能充分利用无线信道的时域相关性及频域相关性,仿真结果显示出了极强的鲁棒性;此外该算法具有的无记忆性,在线训练,递归更新等特点,使其可以低成本硬件实现。This paper proposes a recurvsive least square algorithm combined with regularization and timeweighted techonlogy for on-line training of the radial basis function neural network （RBFNN） so as to solve the problem of wireless channel tracking in orthogonal frequency-division multiplexing （OFDM） system. The approach is able to fully utilize the time-domain correlation information as well as the frequency-domain correlation information of the wireless channel, and is proved to be robust by simulation results. Besides, the algorithm has the property of memoryless, on-line training, recursive updating so that it can be implemented with very low hardware cost.

关 键 词：信道跟踪 径向基函数 OFDM 递归最小二乘 

分 类 号：TN929.533[电子电信—通信与信息系统]