Abstract : Statistical signal processing functions such as signal detection, estimating, and identification play a key role in the development of effective communications, radar, and sonar systems. For example, advanced statistical methods are emerging as being particularly important in digital communications systems operating in channels corrupted by interference from such phenomena as multiple-access noise, intentional jamming, and impulsive noise sources. Conventional demodulation methods, such as coherent matched filtering, often suffer serious performance degradation when subjected to interference of these types; however, this degradation can frequently be eliminated through the use of more sophisticated signal processing techniques. A central issue in the design of effective signal processing procedures for system operating in channels such as those noted above is that of channel identification. Although certain aspects of channel identification have been studied extensively, one area in which there is a pressing need for further research is that of identification of impulsive channels. Communication systems are seldom interfered with by white Gaussian noise alone, yet receiving systems in general use are those which are optimum for white Gaussian noise.