A Novel Analytical Model for Provisioning QoS in Body Area Sensor Networks

Abstract

Wireless Body Area Network (WBAN) has been an active area of research over the past few years due to its tremendous benefits particularly related to healthcare systems. The available research to evolve the QoS in WBAN is immature due to lack of sufficient methodology for modeling the behaviour of different kinds of traffic being generated from different kinds of events. It has been clearly demonstrated that traffic found in multimedia sensor nodes being used in WBAN is having bursty nature and cannot be modeled by using Poisson traffic distributions. However, most of the current available literature of traffic modeling related to Multimedia Wireless Sensor Networks (MWSNs) is based on Poisson distributions. To eliminate these kinds of performance evaluation limitations in MWSNs especially in time critical applications, this study proposes a novel analytical framework that relies on a traffic model resembling to an ON/OFF process. Proposed model exhibit self-similar behaviour and is capable to handle long range dependent traffic patterns. For providing enhanced QoS, proposed model deals with various traffic classes that has been judged in the current study through G/M/1 queuing system with a distinct scheduling strategy called as Low Latency Queuing (LLQ) to extract QoS performance metrics such as delay, queue length, throughput and packet loss rate (PLR). We also simulate the behaviour of traffic to further validate the proposed analytical framework.