OPTIMAL RECEIPT OF MULTIPOSITION SIGNALS USING RANKING METHODS IN CONDITIONS OF PRIOR UNCERTAINTY

Abstract

The article proposes a solution to the problem of signal detection in a priori uncertain conditions and develops a signal detection algorithm against the background of interference. It was established that the flexibility of the ranking procedure makes it possible to solve various problems of signal detection in conditions of a priori uncertainty. The properties of ranking algorithms of multi-position OFDM signals are studied. The unique characteristics of rank compared to other types of nonparametric transformations are analyzed. The selection criterion is not only the study of the given invariant properties of the level of implausible notifications about the type of propagation, but also, which is very important, the maximum possible preservation of information about the signal, which allows for almost complete recovery of the original information, that is, high efficiency of signal detection. Currently, there are no sufficiently effective and acceptable methods of direct estimation of multivariate distribution. In this regard, it is advisable to organize the input sample in such a way that its elements are statistically independent. Ranking procedures make it possible to solve a wide range of signal detection problems under prior uncertainty conditions. A unique feature of the rank in comparison with other types of non-parametric transformations is the ability to almost completely recover the original information. The formation of selective data arrays and their ranking algorithms for multi-position signals with phase-difference modulation of high multiplicity are studied. One of the important problems related to rank processing of information is the question of the place of transition from input data to ranks. The analysis shows that the ranking procedure can be carried out at almost any stage of processing and each of the relevant ranking methods has its advantages and disadvantages.