Time-frequency techniques in wrist pulse signals

Abstract

Wrist pulse signal has been the traditional way of diagnosing human health in India. It is a low frequency, non-stationary signal. There are broadly three techniques of analyzing a signal – time, frequency and time-frequency domain. The individual time and frequency domain representations of the signal do not provide much information on properties of a non-stationary signal. However, the time-frequency distributions of a non-stationary signal overcome these problems and provide more information about the pulse signal. This paper presents the various time-frequency distribution techniques for analyzing non-stationary wrist pulse signals.