The stringent energy constraints of sensor nodes in the Industrial Internet of Things (IIoT) pose a critical challenge to network sustainability. While clustering routing protocols are effective in mitigating energy dissipation, optimal Cluster Head (CH) selection remains an NP-hard optimization problem. Existing metaheuristic approaches often suffer from premature convergence and imbalanced exploration-exploitation capabilities, leading to uneven energy loads and “energy holes.” To address these issues, this paper proposes a novel energy-efficient clustering protocol named CLLDOA-ECP, driven by an enhanced metaheuristic algorithm. First, we introduce the Chaotic Lévy Lens Dream Optimization Algorithm (CLLDOA). This algorithm integrates Logistic Chaotic Mapping for population initialization, Adaptive Lévy Flight for escaping local optima, and Lens Imaging Opposition-Based Learning (LI-OBL) for enhancing convergence precision. The superiority of CLLDOA is rigorously validated against state-of-the-art algorithms (including CSO, GWO, MGO, and KMA) on the CEC2020 benchmark suite, demonstrating exceptional global search efficiency and stability. Subsequently, CLLDOA is applied to the IWSN clustering problem via a specifically designed multi-objective fitness function that balances residual energy, intra-cluster compactness, and distance to the Base Station. Extensive simulation results compare CLLDOA-ECP against six mainstream protocols: LEACH, LEACH-C, EEM-CRP, FOAEAUC-SARP, EBPT-CRA, and VSSLS-SIACR. The experimental data confirms that CLLDOA-ECP significantly optimizes energy distribution and extends the network lifetime by at least 36.6% compared to the benchmark protocols, making it a robust solution for long-term IIoT monitoring tasks.
Sun et al. (Fri,) studied this question.