Recurrence for linearizable switching diffusion with past dependent switching and countable state space

Dang H. Nguyen,George Yin
DOI: https://doi.org/10.48550/arXiv.1710.02888
2017-10-08
Probability
Abstract:This work continues and substantially extends our recent work on switching diffusions with the switching processes that depend on the past states and that take values in a countable state space. That is, the discrete components of the two-component processes take values in a countably infinite set and its switching rates at current time depend on the current value of the continuous component. This paper focuses on recurrence, positive recurrence, and weak stabilization of such systems. In particular, the paper aims to providing more verifiable conditions on recurrence and positive recurrence and related issues. Assuming that the system is linearizable, it provides feasible conditions focusing on the coefficients of the systems for positive recurrence. Then linear feedback controls for weak stabilization are considered. Some illustrative examples are also given.
What problem does this paper attempt to address?