Optimisation of access point deployment model using user-to- ap position vectors in a Wi-Fi IPS

Abstract

Access Point deployment has been a course of major study over the years and numerous researchers have worked upon it. This thesis work focuses on an intelligent deployment of Wireless Access Points (WAP) in an Indoor Wi-Fi Localisation System. Novel in its regard, the work first uses Geometric Dilution of Precision and the Genetic Algorithm to deploy Access Points, followed by a selection of Access Points with the help of RSSI and user-to-AP Position vectors. A series of MATLAB simulations showed that the user-to-AP position vector approach showed an improvement in localisation in the deployment area under study. Further work includes a performance analysis in which simulation parameters are altered to check their effect on system accuracy. Overall, a position vector based selection of Access Points by the users have shown the best results in the CDF plots with better accuracy