Dynamic Channel Reservation Based on Forecast in Cognitive Radio

Channel reservation is a valid technique to reduce dropping probability when spectrum handover happens in cognitive radio. Most of the existing channel reservation schemes focus on proper number of the channels reserved and the number is fixed for a certain system. These static channel reservation schemes bring high blocking probability. In this paper, reusing reserved channels is considered in dynamic channel reservation (DCR) scheme. Secondary users are divided into two classes according to different time-delay-sensitivity in order to use reserved channels dynamically. Renewal theory is used to obtain the forecast results of licensed channels to determine the number of channels reserved. The DCR scheme is analyzed with a Markov chain. Numerical results show that the DCR scheme performs lower blocking probability and higher throughput with dropping probability basically keeping unchanged compared with the schemes based on static channel reservation.