Improved WiFi Based Real-Time Indoor Localization Strategy

Abstract

Accurate mapping and localization of an environment have been improved owing to the advancements in mobile internet technology. However, indoor localization still requires more intelligent algorithms to keep continuous track of a mobile user because of presence of obstacles and satellite’s incapability. The paper presents an indoor wifi based algorithm that combines fingerprint and least square algorithms to track location of a mobile user. The fingerprint data is computed in terms of received signal strength (RSS) acquired from different access points at predefined locations, dynamically. Similarly, the mobile user coordinate is estimated by involving digital filtering process followed by least square technique. The variance in RSS is observed between fingerprint and least square algorithm and applied to Kalman filter for the estimation of weightage value. Thus, the combined mechanism helps to result the value with more accuracy leaving behind low accurate value. The efficiency of the proposed method is evaluated by involving both MATLAB simulation environment covers up to 30 m × 35 m and also hardware resources in real time environment covers up to 13 m × 21 m. The results have showed the accuracy of less than a meter.