Sensor Localization with Lateral Inhibition & Statistical Inference

Abstract

Lateral inhibition plays an important role in biological sensory systems, where sensors are connected such that neighbors exert an inhibitory effect on one another proportional to the distance between them. This report demonstrates lateral inhibition and some of its important properties. For electronic sensors, the method uses radio signal strength as an indication of distance, but is robust to reflection interference because it uses pooling of multiple sensor output. The contribution of this work is a new tool for sensor localization, useful alone or in combination with other methods. A Java¿ application provides the figures shown here, and is available for demonstrating lateral inhibition for different signal strength, sensor layout, and reflectivity values. Computational algorithms are available on request.