Time–frequency feature representation using energy concentration: An overview of recent advances

Signal processing can be found in many applications and its primary goal is to provide underlying information on specific problems for the purpose of decision making. Traditional signal processing approaches assume the stationarity of signals, which in practice is not often satisfied. Hence, time or frequency descriptions alone are insufficient to provide comprehensive information about such signals. On the contrary, time–frequency analysis is more suitable for nonstationary signals. Therefore, this paper provides a status report of feature based signal processing in the time–frequency domain through an overview of recent contributions. The feature considered here is energy concentration. The paper provides an analysis of several classes of feature extractors, i.e., time–frequency representations, and feature classifiers. The results of the literature review indicate that time–frequency domain signal processing using energy concentration as a feature is a very powerful tool and has been utilized in numerous applications. The expectation is that further research and applications of these algorithms will flourish in the near future.