Localization Using GSM Network in Indian Scenario

Abstract

GPS(Global Positioning System) is always considered as a best localization technique. GPS technology uses multiple satellites for finding the exact location. But there are some problems with GPS technology. GPS technology doesn't work in indoor scenario and in similar areas with limited line of sight with GPS satellite. For GPS we need line of sight with the device. Another issue with GPS is specific hardware needed to connect with satellite, which increases the cost. In our project we are trying to get alternative of GPS localization. To achive this we are using cellular network which is more stable and robust network. We are using GSM network to find the location as it has multiple advantages over GPS technology. Using GSM over GPS is beneficial because we don't need any specific hardware basic mobile phones will work. GSM signals will work even in indoor scenario.GPS devices even consume more battery as compare to GSM mobile phones.

In our project we are using GSM fingerprinting algorithm which works in two phase learning phase and prediction phase. In learning phase environment is studied and based on that database is created. In prediction phase based on the database prediction regarding location is done. It works on a simple concept that whenever our GPS is on the algorithm will work in learning phase and will prepare learning database. When our GPS is off and we need to find some location, algorithm will work as prediction phase. It will look into learned database and give best possible matched location.