Stepwise algorithms for improving the accuracy of both deterministic and probabilistic methods in WLAN-based indoor user localisation

The performance of three stepwise algorithms designed to improve the accuracy of WLAN-based indoor localisation methods was determined using a robust simulation technique. The majority of WLAN localisation techniques use the receive signal strength indication to estimate a network user's location using either a deterministic or a probabilistic approach. Previous work has shown that the rms error of these localisation techniques can be over 5 m, which is unacceptably high for indoor applications. Three different stepwise algorithms, two of which are novel, were simulated across several different test bed layouts in order to determine their effectiveness in reducing the error in the position estimate. Of the three algorithms, the constrained movement algorithm offered the best improvement in accuracy with up to a 40% improvement when using deterministic techniques. For probabilistic tracking, there was minimal improvement in the accuracy.