GNB-RPL: Gaussian Naive Bayes for RPL routing protocol in Smart Grid Communications

This study investigates the potential of utilizing the Gaussian Naive Bayes algorithm for enhancing the performance of the Wireless Smart Grid Networks (WSGNs). We have incorporated the Gaussian Naive Bayes algorithm into the widely used Routing Protocol for Low-Power and Lossy Networks (RPL), resulting in an advanced variant named as GNB-RPL. This innovative protocol leverages the Naive Bayes algorithm to optimize routing decisions. Training a Naive Bayes classifier model on a data set of routing metrics enables us to make predictions about the probability of successfully reaching a destination node. Each network node utilizes this classifier to select the route with the highest probability of delivering packets effectively. Our findings demonstrate that GNB-RPL significantly enhances the packet delivery ratio while minimizing end-to-end delay through a comprehensive performance evaluation conducted in a realistic scenario and across different traffic loads. These results show the potential of GNB-RPL as a promising solution for achieving greater efficiency in WSGNs.