MIMO系统中一种鲁棒的叠加导频设计方法

An Robust Superimposed Pilot Scheme in MIMO Systems

与传统的复用式导频相比,叠加导频不占用系统的带宽资源,不会降低信息传输速率,因而受到越来越广泛的关注。基于叠加导频的一阶统计量信道估计算法与最小均方误差算法相比,避免了矩阵求逆运算,大大降低了接收机复杂度,但该算法对信道中的直流干扰非常敏感。本文基于一阶统计量算法,提出了一种优化的叠加导频设计方法,该设计方法在联合最优化信道估计均方误差,接收信干比和峰均功率比的同时,对直流干扰具有很好的鲁棒性。仿真进一步分析了导频参数与信道估计均方误差、接收信干比的关系,比较分析充分表明该设计方法在获得高接收信干比的同时具有良好的鲁棒性。

Comparing with the multiplex training scheme, superimposed pilot will neither consume system bandwidth nor drop the information rate, thus become more and more attractive. The channel estimation algorithm based on one-order statistic can greatly reduce the complexity of receiver by avoiding inversion of matrix,while comparing with the mmse algorithm. ,but the performance could be seriously degrade when a dc-offset exists. In this paper, we propose a optimized superimposed pilot scheme,which not only optimize mse .sinr and papr jointly, but also robust to dc-offset. The relationship of mse and sinr with pilot parameters are analyzed through simulation. Comparision proves that the proposed pilot scheme has both high sinr and good robustness.