Improving localization accuracy of RSS-based lateration methods in indoor environments

To obtain location information, localization methods employing Received Signal Strength (RSS) are attractive since it can reuse the existing wireless infrastructure. Among the large class of localization schemes, RSS-based lateration methods have the advantage of providing closed-form solutions for mathematical analysis. However, the localization accuracy of lateration methods employing RSS is significantly affected by the unpredictable setup in indoors. To improve the applicability of RSS-based lateration methods in indoors, we develop two schemes: regression-based and correlation-based. The regression-based approach uses linear regression to discover a better fit of signal propagation model between RSS and the distance, while the correlation-based approach utilizes the correlation among RSS in local area to obtain more accurate signal propagation. Our results using both simulation as well as real experiments demonstrate that our improved methods outperform the original RSS-based lateration methods significantly, and can achieve comparable or even better accuracy than heuristic-based algorithms such as RADAR and Bayesian Networks, which are prevalent in indoors.