Id3 Based Periodic Sampled Classifier Algorithm for The Enhancement of Routing Protocol in Wireless Sensor Networks

Abstract

Routing protocols for a Wireless Sensor Network (WSN) accomplishes the data dissemination between sensor nodes by preferring the next best node in the routing path. Supervised machine learning comes up with many strategies which can upgrade this next node selection for the range of applications. In this paper, ID3 based next node selection algorithm which uses entropy values for path resumption has been proposed. The simulation study of the algorithm showed that it performs well when compared to k-Nearest Neighbour and Naive Bayes algorithms in terms of evaluation metrics obtained from the confusion matrix and F1-score.