The application of database correlation methods for location detection in GSM networks

A study on the feasibility of applying database correlation methods to locate mobile stations (MSs) within a live GSM cellular network is presented. The mobile station's signal strength values, extracted from the periodic MS measurement reports at the Abis interface, were correlated with the values obtained from a radio network coverage prediction tool, available at the operator's premises, to predict the user's location. The main advantage of this solution is that it uses information that already resides, or is being exchanged, within the network and does not necessitate any additional hardware in the infrastructure. The results obtained were compared with actual road tests, utilizing global positioning systems (GPS), to determine the error of the system. The resulting error correlates well with other studies which do not use dedicated location detection infrastructures.