Modelling an Enhanced Routing Protocol for Wireless Sensor Networks Using Implicit Clustering Technique
The localization of sensor nodes can be a very enabling technology that can help in improving the performance of many algorithms designed for Wireless Sensor Networks (WSNs). This paper is geared towards developing a positioning system that uses received signal strength based on fingerprinting technique. The proposed system models the signal strength distribution received from the sensor nodes using non-parametric or Gaussian distribution. The probabilistic Bayesian technique was employed as the localization algorithm for the basic model. The result obtained shows that there is an improved median error, appropriately 1.0 meter, compared to 2.5 meters for the Nearest Neighbour (NN) algorithm. Implicit clustering technique was used to enhanced the result obtain from the basic model.