End-to-end delay bound analysis in multimedia sensor network using stochastic network calculus

Multimedia sensor network (MSN) has been deployed in many fields owing to its reliability, flexibility and availability of multimedia data. Correspondingly more challenges draw into the QoS guarantee research. In this paper, we establish a comprehensive performance analytical framework for MSN considering both the traffic self-similarity and stochastic service curve of MSN with S-MAC protocol based on stochastic network calculus. With this framework, we derive the m.b.c stochastic arrival curve of Fraction Brownian Motion (FBM) traffic and the stochastic service curve of MSN which adopts sink-tree model. Furthermore, a stochastic end-to-end delay bound for MSN is presented based on the two curves. Moreover we analyze the delay bound qualitatively to get the effect of two key parameters. Finally, numerical experiment and scenario simulation are conducted on Matlab and NS-2 respectively to validate the correctness and tightness of our framework.