Study on the effects of self-similar traffic on the IEEE 802.15.4 wireless sensor networks

Abstract

A significant number of previous studies have shown, however, network traffic exhibited frequently large bursty traffic possesses self-similar properties. For the future applications of wireless sensor networks (WSNs) with large number of cluster structures, such as Internet of Things (IoT) and smart grid, the network traffic should not be assumed as conventional Poisson process. We thus employ ON/OFF traffic source with the duration of heavy-tailed distribution in one or both of the states instead of Poisson traffic to be as the asymptotically self-similar traffic for experimenting on the performance of IEEE 802.15.4 WSNs. In this paper, we will show the impact on the performance of IEEE 802.15.4 WSNs in different traffic sources such as Poisson and Pareto ON/OFF distribution by ns2 simulator. For the Pareto ON/OFF distribution traffic, we demonstrate that the packet delay and throughput appear bursty-like high value in some certain time scales, especially for the low traffic load; and the throughput will be no longer bursty-like while the traffic load increases. Intuitively, the bursty-like high delay may result in loss of some important real-time packets. For the Poisson traffic, both the throughput and packet delay appear non-bursty, especially for the high traffic load.