On Reliable Broadcast in Low Duty-Cycle Wireless Sensor Networks

Broadcast is one of the most fundamental services in wireless sensor networks, where a distinct feature is that sensor nodes may alternate between active and dormant states, so as to conserve energy and extend the network lifetime. Unfortunately, the impact of such cycles has been largely ignored in existing broadcast implementations that adopt the common assumption of all nodes being active all over the time. In this paper, we revisit the broadcast problem with active/dormant cycles. We show strong evidence that conventional broadcast approaches will suffer from severe performance degradation, and, under low duty-cycles, they could easily fail to cover the whole network in an acceptable timeframe. We remodel the broadcast problem in this new context, seeking a balance between efficiency and latency with coverage guarantees. We demonstrate that this problem can be translated into a graph equivalence, and develop a centralized optimal solution. We then extend it to an efficient and scalable distributed implementation. The performance of our solution is evaluated under diverse network configurations. The results suggest that our distributed solution is close to the lower bounds of both time and forwarding costs, and it well resists to the wireless loss with good scalability on the network size and density.