Modeling and Forecasting Secondary User Activity Considering Bulk Arrival and Bulk Departure Traffic Model

Abstract

In this paper, we consider an Erlangian process model for secondary user (SU) traffic in a cognitive radio network (CRN). Unlike prior efforts that focused on Poisson traffic models, the model captures bulk arrival and departure scenarios in a frequency band that is open for opportunistic use by multiple secondary users. Specifically, we develop a Kalman filter based estimation and forecasting strategy for the number of secondary users. Using simulated data, we demonstrate that the proposed approach provides a robust upper bound prediction on the number of secondary users.