FML: Fuzzification with Machine Learning based Parent Node Selection in RPL/6LoWPAN
D. Gopika, Pratham Majumder, Pradeep Kumar · 2020
IPv6 routing protocol for Low Power Lossy Network (RPL) is a standard routing mechanism intended to support the Internet of Things applications. Existing standard RPL mechanisms use an objective function to choose the best parent node within the set of preferred parent nodes, considering a single metric i.e., either hop-counts or expected transmission count for static IoT applications. Still, the mobility metric was unaddressed in the objective function for potential mobility based IoT applications. In this work, we propose a novel Fuzzification with Machine Learning (FML) based algorithm for the selection of best quality parent nodes in Multi Point to Point (MP2P) topology. The simulation result shows 89% of accuracy of prediction with Random Forest (RF) classifier which outperforms SVM (Support Vector Machine) and kNNs (k- Nearest Neighbors).