Asymptotic performance analysis of blind signal copy using fourth-order cumulants
Norman Yuen, B. Friedlander · International Journal of Adaptive Control and Signal Processing · 1996
Blind array processing techniques have been gaining much interest in the last several years. Conventional array processing algorithms such as MUSIC assume that the structural information of the array manifold is known. In blind array processing the structure of the array manifold is assumed to be unknown and needs to be estimated from the available data. In this paper we investigate the performance of blind signal copy algorithms which use fourth-order cumulants to estimate the signals. We derive an asymptotic expression for the upper bound on the quality of the estimated signal in terms of the output signal-to-interference ratio. We also introduce two blind algorithms for signal separation using only fourth-order cumulants: one is asymptotically optimal and the other is suboptimal but significantly simpler. Some insight into the achievable performance of such algorithms is obtained by numerical evaluation of the bound for several test cases of interest and the results are compared with those obtained by simulations using the aforementioned blind algorithms.