Explainable Hybrid ANFIS-Based Multi-Objective Relay Selection Framework for Energy-Efficient Internet of Healthcare Things

Abstract

Efficient, dependable transmission of physiological data is necessary to ensure optimal healthcare monitoring efficiency in Internet of Healthcare Things(IoHT )using IEEE 802.15.6-based networks. However, relay selection remains a major challenge because reliability, latency, throughput, and energy efficiency conflict under variable network conditions. In this light, the paper suggests an Explainable Hybrid ANFIS-based Relay Selection Framework. This novel system combines the Adaptive Neuro-Fuzzy Inference System (ANFIS) with the Non-Dominated Sorting Genetic Algorithm II (NSGA-II) to enable energy-conscious relay selection and dynamic routing. In addition, Shapley Additive exPlanations (SHAP) has been used to improve the interpretability of relay-selection decisions by explaining how specific input feature values contribute to the overall decision. Furthermore, we conduct an ablation study and statistical significance analysis on the effectiveness of different components of the optimization framework. Simulations conducted in IEEE 802.15.6 networking conditions show that the proposed system yields a packet delivery ratio, end-to-end delay, throughput, and residual energy of 98.44 percent, 5.80 milliseconds, 1140 kbps, and 0.860 J, respectively, demonstrating its potential to improve efficiency, latency, throughput, and reliability of healthcare communication using IoHT solutions.