Cross-layer optimization of low power listening MAC protocols for wireless sensor networks

MAC protocols for Wireless Sensor Networks based on channel checking to detect incoming packets (such as Low Power Listening, LPL) require some form of synchronization among sensors, in order to schedule the times in which the sender and the receiver should turn on their radios, and to ensure correct packet delivery. However, while travelling along multihop paths, packets may accumulate delays that force the sensors either to incur packet losses and retransmissions or to readjust the planned scheduling. In both cases this fact impacts on the energy budget of the sensors. In this paper we propose a cross-layer optimization for MAC protocols that use LPL. This optimization takes into account high-level information of the application in order to compute adaptive delays in every sensor along a multihop path, with the goal of adjusting precisely the activity time window of the sensor along the path. We validate our delay-based model by evaluating different scenarios, and we compare it against the LPL model. The simulation results confirm the validity of our approach and demonstrate that a delay-based model can improve the synchronization achieved through the LPL strategy.