MMDEW: An Adaptive Multi-Metric Entropy-Based Objective Function for Stable and Energy-Efficient RPL Routing in Rural Areas

The swift expansion of the Internet of Things (IoT) has accelerated the adoption of Wireless Sensor Networks (WSNs) in various fields, particularly in Low-Power and Lossy Networks (LLNs). The Routing Protocol for LLNs (RPL) is an IPv6-based standard for such environments; however, its default objective function, the Minimum Rank with Hysteresis Objective Function (MRHOF), relies solely on the Expected Transmission Count (ETX), which is often suboptimal in dense and dynamic topologies. This study proposes a novel Multi-Metric Dynamic Entropy Weighting (MMDEW) objective function that integrates four key metrics: ETX, Signal-to-Noise Ratio (SNR), CPU Energy Consumption (EC), and handover frequency, using an adaptive entropy weighting scheme derived from local historical variations. This mechanism enables lightweight, distributed, and self-adaptive parent selections. The proposed approach was evaluated in Contiki-NG under multiple transmission intervals (10, 20, and 30 s) and node densities (12, 15, and 20). Compared with MRHOF, MMDEW reduces handovers by up to 89% and CPU energy consumption by 10–43%, while maintaining a comparable Packet Delivery Ratio (PDR) and End-to-End (E2E) delay. Paired t-test results (p < 0.05, |dz| > 0.8) confirmed MMDEW's significant improvement in route stability and energy efficiency. These findings demonstrate that MMDEW provides a lightweight, distributed, and energy-aware routing solution suitable for long-term rural IoT deployments, such as wildfire early warning systems.