Learning automata and lexical composition method for optimal and load balanced RPL routing in IoT

Abstract

Low power and lossy network, internet of things (IoT) motivates energy-efficient and load-balanced routing in the network layer to extend network lifetime. IoT application scenarios exploit the routing protocol for low-power and lossy networks (RPL) due to the significant potentials. The core components of RPL are the trickle algorithm and objective functions (OF) for creating destination oriented directed acyclic graph (DODAG) and data forwarding. The RPL needs more attention to avoid hotspot problems and unnecessary energy depletion. Most of the existing routing protocols take a single either hop count or ETX, or multiple routing decision metrics. However, the RPL cannot select appropriate link metrics efficiently against the dynamic and lossy environment without considering the relationship between those metrics. Thus, the proposed methodology takes important routing metrics, such as hop count, expected transmission count, and traffic-related metric, and composites the metrics using learning automata and lexical composition method. The special attention on network energy balancing through expected transmission energy (ETT) avoids a hotspot issue and inefficient routing energy. The proposed work supports multiple metrics-based OF with considerable routing overhead by tuning the trickle parameter. Moreover, the proposed work is evaluated to show its advantages over the dynamic and lossy network, IoT.