Modelling And Analysis Of M/G/1 Queuing Systems In Cognitive Radio Networks

Abstract

The drastic change in the present market scenario had brought many revolutionary changes in the cognitive networks by deploying densely populated devices in an efficient manner. The queuing methods are being deployed into the cognitive networks for efficient utilization of the spectrum and to allocate frequencies for the starving users in the queue optimally. Many queuing methods are been proposed for this addressing many solutions among which M/M/1 queuing model is the one which is mostly used. A modified M/G/1 queuing model is proposed with Gamma distribution for service times and a comparative simulation has been done to show that the model provides more stability and better results compared to the existing M/M/1 queuing model that uses exponential distribution for service times. The results have been provided with varying arrival rates of the user and buffer sizes. Also analysed the blocking probability, average waiting time, mean service time and delay wrt varying arrival rate. The results shows that the proposed system model for N users shows a better performance in all the aspects when compared to the existing exponentially distributed service time queuing model..