Abstract-Wireless energy harvesting sensor networks constitute a new paradigm, where the motes deployed in the field are no longer constrained by the limited battery resource, but are able to re-charge themselves through directed electromagnetic energy transfer. The energy sources, whom we call actors, are mobile and move along pre-decided patterns while radiating an appropriate level of energy, sufficient enough to charge the sensors at an acceptable rate. This is the first work that investigates the impact of energy transfer, especially concerning the energy gain in the sensors, the energy spent by the actors, and the overall lifetime in the resulting mobile sensor-actor networks. We propose two event-specific mobility models, where the events occur at the centers of a Voronoi tessellation, and the actors move along either (i) the edges of the Voronoi cells, or (ii) directly from one event center to another. We undertake a comprehensive simulation based study using traces obtained from our experimental energy harvesting circuits powering Mica2 motes. Our results reveal several non-intuitive outcomes, and provide guidelines on which mobility model may be adopted based on the distribution of the events and actors.Index Terms-Wireless sensor and actor networks, energy harvesting, Voronoi, mobility I. INTRODUCTION Powering battery constrained sensors with energy harvesting (EH) has resulted in a new paradigm of long-lived wireless sensor networks (WSN). Such sensors may rely on external, and possibly ambient sources of energy, such as the sun, wind, naturally occurring vibrations, among others, and convert these forms of energy into useful electrical energy that is stored in a capacitor for later use. However, these sources exhibit spatial and temporal variations in the energy that is actually incident on the harvesting circuits, which makes complete dependence on these sources a major concern. Recently, we demonstrated a new technique of powering sensors through electromagnetic radiation in the radio frequency (RF) range [1], which can result in a directed energy transfer. The aim of this paper is to investigate scenarios where the source of energy is mobile, and has power control. Thus, how to move along Event Points (EP) in a WSN while ensuring maximum energy transferred to the sensors in need is the topic of focus in this work.In the rest of this paper we use the term actor to indicate an energy-rich source, which is mobile and can move around in the network. It radiates energy through RF transmissions, which is captured and converted by the on-field sensors connected to energy harvesting circuits.In the architecture considered in this paper, the actors move under different mobility models. They also radiate power at different levels depending upon the distance from the event.
