Anycast Routing Protocol for Forest Monitoring in Rechargeable Wireless Sensor Networks

In rechargeable wireless sensor networks (R-WSNs), sensors have to adjust their duty cycles continuously owing to sporadic availability of energy and operate in a very low duty cycle because of current energy conversion technical limitations. Even though the network lifetime is not the main problem, packet delivery latency is critical because a sensor is in the dormant state most of the time in R-WSNs. These unique characteristics pose a new challenge for routing protocol design over traditional energy-static sensor networks. In this work, we introduce a routing protocol based on anycast communications technology, a novel design to minimize packet delivery latency for forest monitoring in R-WSNs. The key idea is to let a sender choose a closest neighbor to forward packets from forwarding candidates and a nearest sink node to accept the packet from all sinks. For multiple source nodes, we introduce an anycast technology based on Tabu search for each source establishing a minimal E2E delay routing path to reach anyone sinks. Through extensive simulation and experiments, we demonstrate that our anycast scheme based on Tabu search is efficient to provide smallest packets delivery latency in R-WSNs. 1. Introduction Wireless sensor networks (WSNs), a promising technology, have been a topic of much interest to researchers due to their wide-ranging applications. For example, they have been used for military applications, environmental applications, health applications, and home applications [1, 2]. However, a fundamental problem in WSNs is the limited lifetime of sensor nodes. To this end, a significant amount of work has been carried out across the protocol stack to prolong the lifetime of WSNs. Examples of which include energy-efficient medium access control (MAC) protocols [3], duty-cycling strategies [4], energy-efficient routing [5], and topology control mechanisms [6]. Other examples can be found in [7, 8] and references therein. An interesting approach to extend the lifetime of sensor nodes is the use of energy harvesting, which refers to harness energy from the environment or other source. Sensor nodes will equip them with rechargeable technologies that convert sources such as body heat [9], foot strike [10], and finger strokes [11] into electricity, such as a rechargeable battery or an ultracapacitor [12]. The energy from surrounding environment is free, unlimited, and environment friendly. Assuming energy neutral operation [13], a sensor node can operate perpetually if the harvested energy is used at an appropriate rate. Note that a harvesting node is