Introduction

Abstract

The most typical situation for the case of radio communications is the one when there is no a-priory information about direction of input signal. Sometimes, this direction is known till some angular domain. Therefore, there is either complete or partial uncertainty about the spatial structure of the signal. All known methods and algorithms of classical theory of adaptive spatial signal processing (ASSP) are oriented on usage of precise data about spatial structure of the signal. Thus, these classical methods cannot be applied under conditions of a-priory uncertainty about the spatial structure of the signal. This problem was mentioned by many researchers and there are many investigations devoted to this problem solution. There are many attempts to adapt the classical ASSP theory to the case of uncertain input signals. Unfortunately, all known solutions are heuristic ones and they are reduced to some alteration of classical algorithms. The most popular solution is reduced to usage of additional constraints. There are some known heuristic solutions which allow increase of ASSP quality under a-priory uncertainty of the input signal. But there is no solution, which permits to increase the quality of ASSP up to some desirable value if there is an uncommon set of signal-noise situations.