基于子空间相关信道模型的大规模MIMO信道估计

Channel Estimation in Large-Scale MIMO based on Subspace Correlated Channel Model

考虑到大规模MIMO系统的信道向量近似正交以及大量可用自由度,可以对接收信号运用奇异值分解(SVD)理论进行信道估计,并且产生的模糊矩阵用一小段导频就可以完全解决。MIMO系统的性能依赖于传播环境的复杂度和所用天线阵列的性能,更依赖于实际信道所提供的自由度。因此介绍了一种信道模型,该模型中角域划分为有限的信号到达方向,也就是到达角(AOA),并考虑在上行链路中将优于特征值分解(EVD)的SVD运用在AOA相关信道模型上进行信道估计。仿真结果显示,相关信道模型下此估计器的估计精度与基于导频的方法相比可以实现大幅度提高,以达到提高大规模MIMO系统性能的目的。

Since the channel vectors of large-scale MIMO （Multiple-Input Multiple-Output） systems approximate orthogonality and there is plenty of available degree of freedom, the theory of SVD （ Singular Value Decomposition） may be applied to the received signals for estimating the channel, and the generated fuzzy matrix be completely solved by short pilot frequency. The performance of MIMO systems depends on the complexity of communication environment and properties of the adopted antenna arrays, and even on the degree of freedom offered by the physical channels. Thus, the channel model is introduced, in which the angular domain is partitioned into a finite signal direction-of-arrival, i.e. , AOA （ Angle of Arrival）. As SVD is better than EVD on the uplink,it is considered to apply SVD on AOA for estimating the channel. Simulation results indicate that the proposed estimator is significantly increased in estimation accuracy as compared with pilot-based channel method, and this could upgrade the performance of large-scale MIMO system.