Continuous location of moving GSM mobile stations by pattern recognition techniques

Pattern recognition techniques based on hidden Markov models, which are applied to GSM measurement data, have been shown to be appropriate for the determination of the mobile station's location. For computational reasons this is so far limited to a certain number of locations in the road network. Here it is discussed how the number of locations can be increased and how the mobile station can be continuously kept track of once its position has been determined. Splitting the model in smaller parts has been found to be the solution to this problem and it is explained here how this can be treated mathematically.