RPL+: An Improved Parent Selection Strategy for RPL in Wireless Smart Grid Networks

Abstract

Routing protocols play an important role in a Wireless Smart Grid Network (WSGN). The implementation of efficient routing strategies becomes paramount to guarantee the necessary interaction between utilities smart devices like smart meters, and the control centers. This research is focused on improving one of the well known routing protocols for WSGN, the Routing Protocol for Low-Power and Lossy Networks (RPL). More specifically, this paper presents a new parent selection strategy designed to choose the best parent when two or more candidates have the same ranking. The goal is to make better forwarding decisions on the best next-hop node to transmit packets to the destination. The proposed method takes into account the most important routing metrics in this type of network determined by applying Machine Learning (ML) feature importance analysis using Random Forest. The performance evaluation of the proposed strategy shows a significant improvement in terms of Packet Delivery Ratio when comparing to RPL standard implementations.