A binomial distribution approach for the evaluation of indoor positioning systems

Abstract

This work proposes a novel methodology for the evaluation of indoor positioning systems, by adopting a binomial distribution approach during the evaluation process. The positioning accuracy achieved by any Real Time Localisation System (RTLS) is affected by various critical factors, such as the network infrastructure, the implemented positioning algorithms, the environment complexity, the dynamic changes that occur, the device diversity etc. The aim of the suggested methodology is to allow the evaluation of the expected accuracy of a deployed positioning system, under its actual operating conditions. The data collected by the binomial experiment is statistically processed, in order for the evaluator to calculate the strength of his evidence, and finally decide whether the declared accuracy can actually be met or not. The proposed approach, due to its robustness, can be of significant value in device diversity scenarios, organic and dynamic environments.