基于子空间的指数和分解及在信道估计中的应用

Decomposition of exponential sums based on subspace and its application in channel estimation

指数和序列分解问题是一个经典的数学问题,Prony方法和Prony-Kung方法是解决此问题的经典方法。用支撑子空间的观点仔细分析了Prony-Kung方法,在此基础上,理论证明了指数和序列分解问题可以归纳于支撑子空间的确定问题,并对受噪声污染的指数和序列提出了具有低复杂度的Lloyd-like搜索支撑子空间的算法。最后,把以上算法应用于毫米波信道估计,仿真结果表明了此应用的有效性。

The problem of the decomposition of an exponential sum sequence is a classical mathematical problem.Prony method and Prony-Kung method are the classical methods to solve this problem.The Prony-Kung method was analyzed in detail from the point of view of support subspace.On this basis,it was proved theoretically that the problem of the decomposition of an exponential sum sequence could be reduced to the determination of the support subspace,and a Lloyd-like algorithm with low-complexity for searching the support subspace for contaminated exponential sum sequence by noised was proposed.Finally,the above algorithm was applied to estimate millimeter wave channel,and the simulation results show the effectiveness of this application.