A Wlan Fingerprinting Based Indoor Localization Technique Via Artificial Neural Network

WiFi infrastructure provides a great opportunity for indoor localization due to its low cost. Two-dimensional (2D) WLAN based indoor fingerprinting approach has been widely adopted. In 2D positioning scenario, latitude & longitude coordinates of the user are used. But in the real indoor position, condition the position height of mobile device of a user is also an important factor to consider for getting better accuracy in indoor, environment, thus motivates the study into three-dimensional (3D) indoor fingerprinting. To verify mobile height dependency of RSS signal strength on anchor points, a Chi-Square statistical analysis is carried out. RSS based localization is sensitive to various indoor fading effects, which are the main cause of the localization error. A spatial filtering approach is implemented to minimize the fading effects. This paper presents an Indoor positioning system based on 2D and 3D on WiFi Received Signal Strength (RSS) using Artificial Neural Networks (ANN) approach. The obtained results show that using ANN with the proposed method of collecting test data, maximum accuracy of using test data achieved an average distance error of 1.2244m and 1.9065 in 2D & 3D implementation. Future work consists of optimizing the ANN by using genetic algorithms and testing the algorithm over a larger area to test the robustness of the algorithm.