Performance Analysis of a Localized De-synchronization Algorithm for Periodic Data Collection in Wireless Sensor Networks

While energy efficiency is typically considered the major concern in wireless sensor networks (WSNs), many real life applications also require reliability, timeliness and scalability. In such scenarios, Time Division Multiple Access (TDMA) is typically used for data communication, as it avoids collisions and provides predictable latency and minimum energy consumption. However, TDMA requires a strict synchronization among sensor nodes, and exhibits a limited flexibility. In this paper, we propose a de-synchronization algorithm that is able to arrange, dynamically, periodic transmissions from different sensor nodes in a round-robin collision-free style, like in conventional TDMA. Unlike previous de-synchronization schemes, the proposed algorithm is fully localized, i.e. it only relies on local information. In the paper we show how to implement our algorithm on top of IEEE 802.15.4 sensor nodes. We derive and solve an analytical model of the proposed algorithm and use it to analyze its convergence time.