多天线系统符号和载频同步的混合粒子滤波

A Hybrid Particle Filter for Symbol and Carrier Frequency Synchronization in MIMO Systems

为解决多天线系统符号同步和载频同步中的联合参数非线性估计问题,提出一种混合粒子滤波方法.该方法基于最大后验概率准则,在接收端将时延、频率偏置用联合随机粒子表示,结合未知信道参数矢量的卡尔曼滤波估计值,获得联合粒子的重要性权重.同时,采用序列重要性采样和重采样策略来更新粒子权重并保持粒子的多样化,可避免直接在高维信道参数空间上进行重要性采样时效率下降的缺点.该方法在4发2收的多天线基带传输系统上的仿真实验表明,其参数估计精度较高,到达稳态所需的迭代次数较少,并可获得接近理想参数时的误码率性能.

A hybrid particle filter was proposed for the synchronization of symbol and carrier frequency in multiple-input multiple-output wireless communication systems. It is based on maximization of a posterior probability. In the receiver side, propagation delay and carrier frequency offset are represented by a set of random particles. The important weights of particles are calculated by combining the channel parameters, which are estimated by the Kalman filter. Moreover, sequential importance sample and resample are implemented to update the weights and maintain the variety of particles. The approach does not directly sample on multidimensional channel space to overcome the disadvantage of computation efficiency reduction. Numerical results are presented under a four-transmit two-receive wireless communication system. They demonstrate that the estimation accuracy of the approach is high as well as the iterative step is short. The bit error rate performance is comparative to the situation with ideal parameters.