基于FG-PF的双选信道估计与译码迭代算法研究

Channel Estimation and Decoding Iteration Algorithm over Doubly Selective Fading Channels Based on FG-PF

提出一种双选衰落信道下的联合信道估计与译码算法,基于因子图(Factor Graph,FG)与和积算法(Sum-Product Algorithm,SPA),将信道估计器与译码器作为独立的信息处理器融入统一的FG系统模型中,利用不同的消息更新规则,结合粒子滤波算法(Particle Filter,PF)对信道参数进行实时跟踪与估计,实现信道估计器与译码器的迭代消息传递.仿真结果表明,在一定的导频间隔和粒子总数下,算法性能逼近已知信道信息下的系统性能,迭代过程清晰,并有效解决传统算法中由导频数量引起的低频谱利用率问题.

In this paper,a joint channel estimation and decoding algorithm over doubly selective fading channels is proposed.Based on Factor Graph and Sum-Product algorithm,the channel estimator and the decoder,as independent information processors and using different internal/external update rules,are integrated into a unified FG model.Combined with particle filter algorithm,the channel parameters are tracked and estimated in real time to realize the iterative message passing.The simulations show,under certain pilot interval and number of particles,the performance of the algorithm is close to the known channel information with clear iterative process,and effectively solve the low frequency spectrum utilization problem.