Venkatesh T.G.

We present an analytical model for the access during the Service Periods (SP) of the IEEE 802.11ad Hybrid Medium Access Control protocol. As a performance measure of this protocol, we derive the worst case and average delay faced by the SP packets. We show that as the arrival rate of the SP packets increases, the delay increases linearly till a point, beyond which it grows exponentially. Further, we extend the model to variable length of beacon interval, and random allocation of SPs to the nodes. We show how a network designer can do optimal allocation of the SP and the CBAP duration to achieve a tradeoff between SP delay and CBAP throughput. We further extend our analysis for the case of heterogeneous system. Our analytical results are compared with simulation and the results show a good match.

Today, we witness an explosion in wireless networks that are accessed by high-performance mobile devices running a wide variety of demanding applications. The IEEE 802.11ad is a standard targeted at meeting these design requirements. Analytical modeling and performance evaluation is a crucial step needed for the evolution of any standard. In this paper, we model the service periods (SPs) of the IEEE 802.11ad hybrid medium access control as a special case of periodically gated queueing system (PGQS) with Bernoulli scheduling. We first state and prove some basic theorems related to the probability generating functions of the PGQS. Using these results, we derive the average delay and the delay variations (jitter) of packets belonging to SP of IEEE 802.11ad. We show that as the arrival rates increases, the delay increases linearly until the critical arrival rate, and beyond that, it becomes unbounded. Further, a practical scheme for the admission control mechanism of 802.11ad is also presented. To validate the model, analytical results are compared with simulation results.

In this study, the authors present an analytical model for the contention-based access periods (CBAPs) and service periods (SPs) as specified in the medium access control (MAC) layer of IEEE 802.11ad standards. The analytical model for CBAP is based on a two-dimensional Markov chain which captures the hybrid nature of IEEE 802.11ad MAC. The Markov chain is used to obtain important metrics like MAC throughput and average frame service time. The SPs are modelled as a M/G/1 queueing system with vacations. Using this model, the delay experienced by packets is calculated. The accuracy of the analytical models is established by extensive simulation results. A brief insight into optimal allocation of SP and CBAP is also discussed.