In this thesis, large error mitigation algorithm for time based location system in indoor sensor networks has been proposed. The proposed algorithm is based on TDOA Spherical Intersection (SX) method.All transmitters and receivers in TOA geo-location ...
In this thesis, large error mitigation algorithm for time based location system in indoor sensor networks has been proposed. The proposed algorithm is based on TDOA Spherical Intersection (SX) method.All transmitters and receivers in TOA geo-location system have to be precisely synchronized. However such a precise synchronization is not required in TDOA geo-location system. And the SX method has an advantage that the location estimate is produced simply from the solution of closed form equation. But the SX method presents different performances according to the distribution of position of sensors and the mobile. If the sensors are linearly located, estimate can not be obtained or have large errors.As the one of large errors in indoor networks, NLOS errors should be considered, since NLOS errors usually have unknown distribution with a positive mean, therefore cause the dominant errors in location estimation [1].The TDOA-RWGH algorithm proposed in this thesis can be operated in the condition of the same number of minimum required sensors in TOA system by applying modified residuals in TDOA system. The proposed scheme can reduce the NLOS errors and the rate of non-solution caused by the topology of random distributed sensors effectively. Also, by using the modified residuals, the method that classify dually calculated (the distance from origin to mobile) in SX algorithm can be obtained, which is effective for random distributed sensor network