A Novel Kalman Filter Based Trilateration Approach for Indoor Localization Problem

Accurate identification of location of a mobile object in indoor environment is very much important due to its role in location based services. Global positioning system suffers in indoor environment due to poor signal strengths of distant satellite. In indoor environment, positioning information is obtained by processing the received signal strengths communicated between the mobile object and various stationary wireless access points. The noise contaminating the measurements and the propagation delay between receivers and senders make this processing complicated. A Kalman filter can be utilized to handle such intricacies. Accuracy of such Kalman filter based approach is very much depended on initialization of parameters, which is further depended on accurate knowledge of the location map. Associating a Kalman filter to preprocess the measurements of all access points of the location makes the system computationally expensive. The current work investigates a Kalman filter based indoor localization system which avoids the need of any prior knowledge of the environment which is essential in methods such as fingerprinting. Instead of preprocessing the measurements available from all access points, it first uses trilateration based localization algorithm on how many data are available. It then applies one Kalman filter algorithm on the data which found nearest to the object based on the proximity obtained in the previous phase. This makes the system computationally efficient. Simulation results show that, < 1 meter accuracy can be obtained by this technique which is at par with some existing techniques.