USENIX Association 10th International Conference on Autonomic Computing (ICAC ’13) 193 Between Neighbors: Neighbor Discovery Analysis

Abstract—Devices in future Internet of Things (IoT) will be scavenging energy from the ambiance for all their operations. They face challenges in various aspects of network organization and operation due to the nature of ambient energy sources such as, solar insolation, vibration and motion. In this paper we analyze the classical two-way algorithm for neighbor discovery (ND) in an energy harvesting IoT. Through analysis, we outline the parameters that play an important role in ND performance such as node density, duty cycle, beamwidth and energy profile. We also provide simulation results to understand the impact of the energy storage element of energy harvesting devices in the ND process. We demonstrate that there exist trade-offs in choices for antenna beamwidth and node duty cycle, given node density and energy arrival rate. We show that the variations in energy availability impact ND performance. We also demonstrate that the right size of the storage buffer can smooth the effects of energy variability. I