PSGWO: An Energy-efficient Framework in IoT Based on Swarm Intelligence

Abstract

Background: Internet-of-things (IoT) has been developed for use in a variety of fields in recent years. The IoT network is embedded with numerous sensors that can sense data directly from the environment. The network's sensing components function as sources, observing environmental occurrences and sending important data to the appropriate data centers. When the sensors detect the stated development, they send the data to a central station. On the other hand, sensors have limited processing, energy, transmission, and memory capacities, which might have a detrimental influence on the system. Objectives: We have suggested an energy-efficient framework based on Swarm Intelligence in IoT. The idea behind using Swarm Intelligence is the probabilistic-based global search phenomena that suit well for IoT networks because of the randomization of nodes. Our framework considers the prominent metaheuristic concepts responsible for the overall performance of the IoT network. Our current research is based on lowering sensor energy consumption in IoT networks, resulting in a longer network lifetime. Methods: This study selects the most appropriate potential node in the IoT network to make it energy-efficient. It suggests a technique combining PSO's exploitation capabilities with the GWO's exploration capabilities to avoid local minima problems and convergence issues. The proposed method PSGWO is compared with the traditional PSO, GWO, Hybrid WSO-SA, and HABC-MBOA algorithms based on several performance metrics in our research study. Results: The results of our tests reveal that this hybrid strategy beats all other ways tested, and the energy consumption rate of the proposed framework is decreased by 23.8% in the case of PSO, 20.2% in the case of GWO, 31.5% in the case of hybrid WSO-SA, and 29.6% in the case of HABC-MBOA, respectively. Conclusion: In this study, several performance parameters, including energy consumption, network lifetime, live nodes, temperature, and throughput, are taken into account to choose the best potential node for the IoT network. Using various simulations, the performance of the proposed algorithm was evaluated and compared to the metaheuristic techniques. Moreover, PSGWO is found to be improved, and the energy consumption rate is decreased.