STOCHASTIC ANALYSIS OF GSB PROCESS

摘要

We present a modification (and partly a generalization) of STOPBREAK process, which is the stochastic model of time series with permanent, emphatic fluctuations. The threshold regime of the process is obtained by using, so called, noise indicator. Now, the model, named the General Split-BREAK (GSB) process, is investigated in terms of its basic stochastic properties. We analyze some necessary and sufficient conditions of the existence of stationary GSB process, and we describe its correlation structure. Also, we define the sequence of the increments of the GSB process, named Split-MA process. Besides the standard investigation of stochastic properties of this process, we also give the conditions of its invertibility.