WiFi-RTT Indoor Positioning Using Particle, Genetic and Grid Filters with RSSI-Based Outlier Detection

WiFi has vast infrastructure presence making it an ideal candidate for mobile indoor positioning. WiFi Fine Time Measurement (FTM), is a WiFi protocol that enables the time of flight (ToF) of a WiFi signal to be determined; referred to as WiFi Round Trip Timing (RTT). Providing a ToF based protocol has allowed ToF based positioning algorithms to be applied to WiFi signals which could provide an improvement over the current RSSI-fingerprinting state of the art. Non line of sight (NLOS) reception and multipath interference degrade WiFi RTT accuracy. The research in this paper explores the accuracy of WiFi RTT positioning in a variety of indoor environments by utilising filtering techniques and RSSI-based outlier detection. Four positioning algorithms are explored: Least squares, a particle filter, a genetic filter and a grid filter. 67% of trials resulted in sub-metre accuracy and 90.5% of trials had a RMSE below 2m, the accuracy was worst in environments with NLOS conditions where 38% of trials resulted in sub-metre accuracy whereas for environments with complete LOS conditions 95.2% of trials resulted in sub-metre accuracy. A method to mitigate NLOS error is RSSI-based outlier detection, this method detects anomalies between the RSSI and the measured RTT range and de-weights anomalous signals during filtering. This outlier detection performed well in environments with NLOS conditions, at its best providing an average improvement of 41.3% over no outlier detection across all algorithms in an environment. The Genetic Filter performed best overall with a mean improvement of 49.2% when compared to least squares, the particle filter performed achieved an average of 38%. For the particle filter, this can be attributed to poorer mitigation of particle degeneracy. The genetic filter was also the only algorithm to provide a performance improvement over least squares in all environments.