Feasibility analysis of vehicular dynamic spectrum access via queueing theory model

Abstract

In this article, we present a feasibility analysis for performing vehicular dynamic spectrum access across vacant television channels via a queueing theory approach. In this study, we model the vacant TV channels as a multiserver, multipriority, preemptive queue. Queueing theory has been extensively employed in the open literature to model multiple access architectures as well as analyze network performance in both wired and wireless communication frameworks. On the other hand, to the best of the authors' knowledge, not many researchers have tried to resort to a queueing theory approach to analyze dynamic spectrum access networks on a system level, including the specific case of vehicular dynamic spectrum access (VDSA). Leveraging previously reported quantitative measurements obtained from a wireless spectrum measurement campaign conducted along a major interstate highway (I-90) located in Massachusetts, we modeled vacant TV channels as a multiserver queueing system in which available servers represent vacant TV channels. The servers become unavailable from the perspective of a vehicle in a time/location-varying fashion such that they represent spatially occupied TV channels. Both M/M/m and M/G/m models are employed to evaluate the probability that a vehicle finds all channels busy, as well as to derive the expected waiting times. We also consider cases where there are multiple priority classes of service requests such as a channel request by a first--responder vehicle.