The Analysis of Synchronous All-to-All Communication Protocols for Wireless Systems

Many applications running over low-power and lossy wireless networks and wireless sensor networks (WSNs) rely heavily on a number of all-to-all communication primitives for services such as data aggregation, voting and consensus. Starting with the Chaos system, synchronous transmission-based broadcasting gossip protocols are now recognized as a technique for enabling efficient all-to-all communications in WSNs. However, despite their effectiveness, there has been relatively little analysis of this class of synchronous broadcasting gossip protocols (SBGPs). In this paper, we address this void by providing a basic theoretical framework for analysis SBGPs. Based on our derived theoretical results and previous experimental measurements, we show that the key for better performance is to increase the network connectivity as much as possible while limiting the number of concurrent transmitters. As a proof of the concept, we propose a multi-radio approach of the SBGP to achieve this purpose. We compare four multi-radio schemes of SBGP with a single radio SBGP through simulation and result has shown the convergence latency can be reduced up to 42% by utilizing multiple radios.