Temporal analysis of stationary Markov a-sub-Gaussian noise

In warm shallow waters, snapping shrimp noise is the dominant source of ambient noise at medium-to-high frequencies. The noise process is impulsive and exhibits memory. The latter property causes outliers to cluster together, thus resulting in bursty impulsive noise. When tuned to snapping shrimp data, the stationary α-sub-Gaussian noise with memory order m (αSGN(m)) model tracks the former's temporal amplitude statistics very well. In comparison to contemporary impulsive noise models, αSGN(m) offers a far more realistic model for snapping shrimp noise. To develop a more deeper understanding of αSGN(m), we perform an impulsive event analysis on its realizations. This is accomplished by initially mapping impulsive data to a point process in time. The resulting time-series is then analyzed via first-order interval and counting analysis. We compare our results with those of snapping shrimp noise and highlight the pros and cons of adopting the αSGN(m) model. Moreover, our results also offer us a reliable way to tune the order m of αSGN(m).