基于确定性观测的压缩感知MIMO信道参数反馈

在实际资源受限（带宽受限或功率受限）的无线通信系统中，多径信道具有很强的稀疏特性，如何利用这一特点更加高效地将接收端获得的多径信道状态信息（CSI）进行压缩、反馈，是目前信道状态信息反馈技术的研究热点。针对现有多入多出（MIMO）信道状态信息反馈方法中随机观测矩阵需要较大存储空间的问题，引入了确定性Golay测量矩阵，提出了一种基于确定性观测的压缩感知MIMO多径信道参数反馈方法。在接收端对由信道估计得出的信道状态信息利用确定性Golay测量矩阵进行观测，将较少数目的观测值反馈到发送端，在发送端通过重构算法恢复出完全信道状态信息。仿真实验表明，与随机观测相比，采用确定性Golay观测矩阵的方法虽然需要的观测值数目会有所增加，但所需存储空间远小于随机观测矩阵，且利用确定性观测反馈信道状态信息的重构性能与随机观测矩阵相当。

Compressed sensing MIMO channel parameters feedback based on deterministic observation

In bandwidth-limited or power-limited wireless communication system,how to more effectively compress and feedback the multipath channel state information (CSI) obtained from the receiver is a hot topic in CSI feedback field. In consideration of existing feedback methods of CSI in multiple-input multiple-output(MIMO) system,large storage space is indispensable for compressed sensing with random observation.This paper proposes a new CSI feedback method for MIMO multipath channel based on the deterministic observation of compressed sensing. CSI estimated by the receiver is observed by deterministic Golay measurement matrix,and a part of observation values are fed back to the transmitter. Then,CSI can be recovered completely through reconstruction algorithm at transmitter. Simulation results show that in comparison with random observation matrix,the Golay observation matrix needs more observations slightly,but the required memory is much less. And CSI can be reconstructed as good as random observation matrix. Using deterministic Golay observation matrix can greatly reduce system overhead and improve spectrum efficiency and system performance.