On analysis of exponentially decaying pulse signals using stochastic volatility model. Part II : student-t distribution

Abstract

The authors have recently demonstrated how the stochastic volatility model incorporating the exponential power distribution can be used to retrieve the instant of initiation of an exponentially decaying pulse and its decay constants in the presence of background noises. In the present study, the Student- t distribution, which can be expressed in a two-stage scale mixtures representation, is adopted in the stochastic volatility model. It is found that the corresponding performance is comparable to that for the case of the exponential power distribution when the signal-to-noise ratio is larger than or equal to 3 dB. The performance deteriorates quickly when the signal-to-noise ratio drops below 0 dB.