Queueing Delay Performance of Schedulers in Centralized Sensing Cognitive Radio Networks

Abstract

Wireless communication has become cost-effective and it can be established very rapidly in order to bring broadband access to under-served rural as well as urban area users. The underutilized spectrum has enormous, untapped, public capacity to provide high-speed and pervasive broadband connectivity. The wireless Radio Frequency (RF) spectrum is highly congested due to the continuous increase in the demand for high data rate services. Cognitive Radio (CR) allows secondary/unlicensed users to use a part/slice of a licensed spectrum, referred to as a channel, whenever the primary/licensed user is not transmitting in that channel during a period. In order to find the primary–free channels in the RF spectrum, CR performs spectrum sensing before transmitting its signal. When there are multiple secondary users trying to sense, find and access the available free channels it becomes very challenging. Scheduling resolves the problem by the use of queues. The behaviour of queues in terms of Queuing delay performance of Schedulers in Multiuser Cognitive Radio Networks when there is a Centralized Coordinator performs sensing and scheduling, has been studied. Three different scheduling policies are considered, namely Maximum Gain, Maximum Weight and Proportionally Fair scheduling. When the network is heavily loaded, each policy presents varied Queuing Delay Performance. The cases when average Signal-to-Noise Ratio of users is considered to be unequal have been analysed with varying normalized sensing duration. Queuing delay for various arrival rates of users is plotted and the performances have been compared and discussed. From the results, it can be concluded that the Maximum Weight Scheduler performs better than others do.