Exploring opportunistic access techniques using stochastic models: Dynamic spectrum access without sensing

In this paper we explore the use of stochastic approaches for opportunistic access based on models of primary user activity. We focus in particular on a scenario where the secondary user does not perform traditional sensing at all, but relies only on statistical information of primary transmissions. Using techniques from the theory of alternating renewal processes we show that, for different primary user constraints, simple statistical transmission schemes exist based on selecting sufficiently low transmission probability for the secondary user. We describe the general framework for deriving the required results, and give a detailed case study assuming that the primary transmitter activity patterns follow exponential distributions. The results for the exponential case show that limited but stable capacity can be achieved by the secondary user for any primary activity level. We also discuss the possible usage scenarios for such statistical approaches, focusing on situations in which traditional sensing is challenging due to high level of dynamics in the environment.