T-AAD: Lightweight traffic auto-adaptations for low-power MAC protocols

After many successful deployments, Wireless Sensor Networks (WSNs) now appear open for a paradigm shift in traffic resulting from applications. Consequently to the traditional event or time-driven a priori known traffic patterns, those networks face occasional, bursty and unanticipated multi-hop data transmissions. This specific case is rarely addressed by preamble-sampling Medium Access Control (MAC) protocols. Such homogeneously configured solutions avoid network disconnections and isolated nodes, yet lead to long periods of channel occupancy, increased delays and energy-consumption. We here present T-AAD, a Traffic Auto-ADaptive mechanism, specifically designed to address the previously reported phenomenon through the introduction of heterogeneous node configurations in the network. T-AAD dynamically and locally adapts the MAC configuration depending on the actual and expected traffic load, without endangering network connectivity nor overall network performances. T-AAD is therefore compliant with Low Power Listening (LPL) based MAC protocols. In this paper, we evaluate our proposed mechanism and we demonstrate that it provides significant gains in terms of delay and energy consumption, when compared to respective research work available in the literature (e.g. MaxMAC, AADCC).