MIMO-OFDM系统中反馈降低算法

Feedback reduction algorithms for MIMO-OFDM systems

本文针对MIMO-OFDM波束赋形系统中基于子载波递归反馈编码算法反馈负载过大和基于簇递归反馈编码算法出现跟踪错误传播问题,提出一种新的周期性递归反馈编码算法。该方法利用簇波束向量之间的冗余相关性进一步降低反馈量。同时,通过周期性地搜索初始码书避免了跟踪错误的传播问题。另外,还提出了一种修订分簇反馈降低方法,接收端只当码字发生变化时才反馈最优码字及其子载波索引给发送端,否则不反馈任何信息。仿真结果表明这两种方法可以有效地降低反馈量,并且保证了系统性能。

The recursive coding method based on subcarriers for multiple input multiple output orthogonal frequency division（MIMO-OFDM） systems has large feedback overhead,and the clustering based recursive feedback coding approach appears error propagation problems.Therefore,a novel feedback reduction method based on recursive coding is proposed.Exploiting the residual frequency correlation between the cluster beamforming vectors,this method reduces feedback load more efficiently.Meanwhile,the proposed approach avoids error propagation by searching initial codebook periodically.We also propose a new modified clustering feedback reduction method,which will inform transmitter the changed codewords and their indexes.Simulation results verify the significantly reduced signaling overhead with good system performance.