Abstract
Localisation is a challenging issue in wireless sensor network. This paper describes a neural network and fuzzy logic-based approach for localisation in wireless sensor network (WSN). The received signal strength indicators (RSSIs) of some anchor nodes are used as basic parameter to estimate the location of sensor nodes. Using fuzzy logic the RSSI values of anchor nodes are categorised into some predefined regions and RSSI patterns are generated using fuzzy inference rules. These patterns are then used as input to a trained neural network (NN) for estimating a proximity factor of sensor nodes which in turn is used to calculate their positions. The RSSI patterns are used to find out the weighted position of the anchor nodes which when divided by the proximity factor gives the estimated position of the sensor nodes. A modified back propagation method is used to train the neural network. Proposed model is tested using network simulator NS2 and result shows accuracy up to 95%.
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call