Who Face Me? Neighbor Identification With LED Camera Communication

Abstract

Neighbor discovery is a fundamental task to support many other services in wireless multihop networks (WMN). Most existing related methods in WMN rely heavily on the information of radio waves. To extend the way to the neighbor discovery, this paper introduces another interesting way, visual light communication, to explore its property of linear transmission. We apply light emitting diode (LED) array and camera and present a novel communication system, named LED array to camera system (LC). This paper also designs a novel protocol, named LED to camera protocol (LCP), for it. Equipped with an LC, each node in a WMN can determine the precise direction and distance of its neighbors and recognize their identities. Furthermore, this paper also develops a method to infer the topology of the whole network. We design the hardware for LC and conduct extensive experiments to implement the protocol LCP. The average latency of the neighbor discovery is measured and can be as small as 1.087 seconds. LCP can achieve centimeter-level accuracy in distance and direction to averages of 0.37 cm and 1.67 degrees in these real experiments. The relative accuracies in distance and direction measurement are 99.11% and 88.92% on average, respectively. The simulation of the topology inference is also performed to show the feasibility and accuracy with the local distance and direction information.