Performance Evaluation of Refinement Method in Indoor Localization

Indoor localization system becomes an interesting topic for the researcher, especially when we estimate the position of object based on RSSI. RSSI depends on the fading or multipath in indoor so that the estimation will be not accurate anymore. Some popular methods such as trilateration and multilateration had been tried to detect the relative position of user using the distance from anchor node to mobile node. To reduce the error from error geometric effect in positioning, GDOP can be implemented in 2 dimensions of indoor localization. We try to combine GDOP with trilateration and multilateration, and evaluate the system using four methods. By using GDOP and multilateration, accuracy of positioning can be increased. The accuracy of trilateration and GDOP combination decreases by 2% than compared to trilateration. But when we estimated the position using multilateration, the accuracy increased by 9.5%. Then, we combined GDOP with multilateration, and the accuracy reached by 13%. So that, refinement method such as combined of multilateration and GDOP can improve the accuracy of estimation position than in comparation to trilateration method.