The role of statistics in signal processing-a brief review and some emerging trends

Abstract

Statistical signal processing plays a crucial role in the design of many modem engineering systems. The availability of high performance, high speed DSP (digital signal processing) devices has seen this role increase by leaps and bounds. This paper takes up an overview study of the developments in this field with special reference to the last two decades as well as currently emerging trends. After a quick review of the Wiener and Kalman filtering theories, which served as major precursors to modern statistical signal processing, we address the important developments in adaptive filtering and modelling- a subject which has received a lot of attention in the eighties, and which continues to make strides to this day. Another area of considerable interest in the recent literature is concerned with the problems of harmonic retrieval and direction-of-arrival (DOA) estimation of plane waves from sensor array data. Finally, the recent emergence of ''higher order statistics'' (HOS) like cumulants and polyspectra as important tools in signal processing is highlighted.