Wi-Fi based indoor localization using trilateration and fingerprinting methods

Nowadays, mobile devices such as personal computers and smartphones are emerging as a major key in today’s computing platforms for indoor object localization systems due to the object localization in indoor areas strongly suffers from limitation of using GNSS (Global navigation satellite system) systems due to low satellite availability and high signal attenuation. During the last decade, many researchers have developed indoor localization systems which are the process of obtaining user or device location through mobile devices using Wireless Fidelity (Wi-Fi) network signals with promising results and acceptable performance. In these Wi-Fi based localization systems, indoor positioning relies on different types of measurements including Time-Of-Arrival (TOA), Time-Difference-Of-Arrival (TDOA), Angle-Of-Arrival (AOA), and Received Signal Strength (RSS) of Wi-Fi signal. In this paper, the techniques and algorithms that used for the RSS-based localization such as Trilateration and Fingerprinting which depend on the RSS from the access point (WI-FI). Using
Received Signal Strength Ranging approach in the Trilateration method which depends on database that contains path-loss-exponent and shadowing parameter that differ according to the environment, solving the equation using different Access Points (APs) at least 3 APs, the no of the APs and there locations were varied to get the best accuracy which depends on the horizontal dilution of position (HDOP). At the Fingerprinting method depends on matching the recorded offline RSS from nearby access points (AP) to the online RSS received by the user on the move is reviewed. A comparison
of location fingerprinting methods involving deterministic method (k-nearest neighbor method and weighted k-nearest neighbor method), probabilistic methods by estimation of likelihood functions with several approaches (non-parametric and parametric)are also explained. The performance parameters of this study include the two-dimensional root mean square error (2D-rms) which measures the localization accuracy. Moreover, the effect of increasing/decreasing the number of APs on the system accuracy is also discussed. The aim of this paper is to announce which method can provide better performance than the others and under what conditions.