LAUSR

Abstract Mobile sensor networks (MSNs) have become a research hotspot recently and lead to a wide demand for wireless communication-based applications such as Internet of Things (IoTs), vehicular networks and mobile social networks, etc. However, due to the power limitation of battery and the requirement of real time performance, it is critical to design an effective and efficient neighbor discovery protocol with minimal energy consumption and discovery latency. Prior discovery protocols are essentially static, in which the passive discovery can be achieved when a pair of neighboring nodes are active at the same time according to their predefined schedules. In this paper, we design a new neighbor discovery model, in which the beacons are separated from the active slots and the broadcast of beacons can be dynamically adjusted to accelerate the discovery. Based on the model, we propose a proactive wakeup based energy-efficient neighbor discovery protocol called PWEND. The PWEND is applicable to both the slot-aligned and slot-unaligned scenarios, and the worst-case discovery latency is obtained. Furthermore, we extend the PWEND to make it applicable to the asymmetric scenario, where different nodes are with different duty cycles. We implement and evaluate the PWEND protocol by state-based simulations and the simulation results show that PWEND protocol outperforms other existing ones.