RSS based Wi-Fi positioning method using multi layer neural networks

In wireless communication system, indoor localization is necessary to provide the exact location of a radio receiver. Estimation of receiver position is difficult in indoor environment due to severe attenuation of radio signals. There are different indoor positioning algorithms like Time of Arrival (TOA), Angle of Arrival (AOA), Triangulation and Fingerprinting. In order to overcome the problem of nonlinear relationship, neural networks can be used. Different Neural networks are used to estimate the location of receiver in indoor positioning. In this paper Levenberg-Marquardt (LM) algorithm in neural network is employed to find the accurate location of receiver device in KL Education Foundation, Guntur, Andhra Pradesh. The estimated locations of receiver are drafted as a path and compared with theoretical values which are represented as coordinates of X and Y. It is evident that from experimental results Levenberg-Marquardt (LM) algorithm has better performance than other existing algorithms. The outcome of this work would be useful wi-fi finger printing applications.