High-accuracy localization via measurements of RSSIs and LED light angles for low-cost WMSNs

Abstract

Wireless multimedia sensor networks (WMSNs), constructed from multiple sensor nodes with cameras, are expected to collect visual data over wide ranges. Existing studies have attempted to estimate the positions of nodes using received signal strength indicators (RSSIs). To improve the accuracy of the estimations of node positions, we propose a new localization method using both RSSI and images sensed by camera sensor nodes for WMSNs. Camera sensor nodes, with positions that are unknown measure the angles of LED light signals emitted by anchor nodes and their RSSI values, Using the RSSI values, nodes can calculate the distances from the anchor nodes. In addition, using the plural angle values, nodes can calculate their own position with high accuracy. We built a small WMSN test bed and measured the average estimation errors. We confirmed that the average estimation errors of a method using only the RSSI values and our proposed method are 1.07 m and 0.29 m, respectively. Therefore, the proposed method could reduce the estimation error by approximately 73%.