Improvement of Power Efficiency of Sensor Network Using Ant Colony Optimization and K-Means Clustering Technique

• Nsikan L. AKPAN

  Author

• Akaniyene OBOT

  Author

• Olusegun A. AFOLABI

  Author

Energy-constrained routing remains a central limitation in wireless sensor networks (WSNs), because repeated transmission, poor cluster-head selection, and unreliable paths accelerate battery depletion and fragment sensing coverage. This paper presents a K-means-guided adaptive multi-objective ant colony optimization (KMA-MOACO) framework that combines logical node clustering, residual-energy-based cluster-head rotation, multi-pheromone route construction, and Pareto-dominance-based selection. The routing problem is formulated to minimize communication energy and end-to-end delay while maximizing packet-delivery reliability, subject to node-energy, transmission-range, link-quality, and delay constraints. A first-order radio model is used in a 100 m × 100 m simulation field with 50–150 stationary nodes; the reported comparative case contains 100 nodes initialized with 2 J, periodic 64-byte packets, and 10–30 ants per routing round. KMA-MOACO is evaluated against a conventional ACO routing scheme and a generic multi-objective ACO implementation. In the reported representative case, KMA-MOACO reduced average energy consumption to 0.5676 J and average end-to-end delay to 0.3585 s, corresponding to approximately 75% reductions relative to both comparators. It achieved a packet-delivery ratio and normalized throughput of 0.92, full node reachability, an energy-delay product of 0.2035, and an energy-reliability product of 0.6170. The generic multi-objective ACO retained the longest raw network lifetime and the largest hypervolume, whereas the proposed method provided the strongest overall compromise between energy, latency, connectivity, and reliability. The findings support adaptive clustering and Pareto-guided pheromone control as a practical direction for resource-constrained WSNs, while also identifying the need for repeated-seed statistical validation and hardware-in-the-loop assessment.

• 17_Akpan et al. (2026)

Copyright (c) 2026 Nsikan L. AKPAN, Akaniyene OBOT, Olusegun A. AFOLABI (Author)

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.