Applying Blind Signal Processing Methods in Data-Transmission Systems

Abstract

The problem of increasing the noise-proof feature of communication systems, along with increasing the communication channel capacity, is currently important. In many areas of wireless engineering (in radiolocation and communications for example), it becomes necessary to receive and process signals, some characteristics of which (for example, amplitude, frequency, or initial phase) are unknown. Methods of statistical radio engineering, as well as algorithms for blind signal processing, can be used to eliminate a priori uncertainty. The problem of comparative analysis of some common algorithms for blind signal processing in the tasks of separating radio signals from mixtures with interference and noise was considered. In the study, the algorithms JADE, FastICA, SOBI and AMUSE were considered. Amplitude-, frequency- and phase-shift keyed signals against the background of additive noise, impulse noise and cochannel noise were considered as useful signals. Based on the results of simulation modeling, the effectiveness of these algorithms is compared when distinguishing between several types of signals in the presence of various types of interference. The above research results can be useful when choosing a blind processing algorithm for subsequent practical application.