FUSION OF MULTILAYER PERCEPTRON AND KALMAN FILTER FOR INDOOR OBJECT TRACKING
Abstract
The indoor object tracking using received signal strength (RSS) or received signal strength indicator (RSSI) measurements in wireless sensor network (WSN) is of utmost important in the context of many location based services (LBS). Without the knowledge of location, the measurements obtained with WSN are of no use. The trilateration is widely used technique to get location updates of target based on RSSI measurements from WSN. However it suffers with high localization errors due to fluctuating nature of RSSI. This paper presents a range free multilayer perceptron (MLP) and kalman filtering (KF) based algorithm named, MLP+KF. The performance of MLP+KF algorithm is evaluated using simulated RSSIs and are compared against trilateration based estimation and MLP based estimation. The simulation results reveal that proposed MLP+KF algorithm show very high localization accuracy as compared to rest of the two approaches.