Adaptive algorithm for detection of signals in an impulsive interference environment

Abstract

Impulsive interferences are frequently long relative to signal duration. This situation is observed, for example, in a telephone channel used for medium and high-speed data communication. In this paper a novel algorithm for signal detection under conditions of long impulsive interferences is proposed. The approach does not require a priori knowledge of the impulse noise statistics. It is based on the representation of the impulse noise using a linear prediction model. Performance of I he proposed algorithm is investigated using computer simulation. It is shown that the algorithm provides a greatly reduced error rate as compared with the conventional non-linear receiver. Its performance is virtually independent of the power of impulse noise.