Estimation of Time-Varying Channel State Transition Probabilities for Cognitive Radio Systems by means of Particle Swarm Optimization

Abstract

In this study, Particle Swarm Optimization is applied for the estimation of the channel state transition probabilities. Unlike most other studies, where the channel state transition probabilities are assumed to be known and/or constant, in this study, these values are realistically considered to be time-varying parameters, which are unknown to the secondary users of the cognitive radio systems. The results of this study demonstrate the following: without any a priori information about the channel characteristics, even in a very transient environment, it is quite possible to achieve reasonable estimates of channel state transition probabilities with a practical and simple implementation.