A novel hybrid metaheuristic method for efficient decentralized IoT network layouts

Abstract

This paper introduces a Hybrid Genetic Particle Swarm Optimization (HGPSO) method focusing on optimal and efficient sensor deployment in Wireless Sensor Networks (WSNs) and Decentralized IoT (DIoT) networks. Effective sensor placement in these networks necessitates the simultaneous optimization of numerous conflicting goals, such as maximizing coverage, ensuring connectivity, minimizing redundancy, and improving energy economy. Traditional optimization techniques and single metaheuristic algorithms frequently encounter these difficulties, demonstrating premature convergence or inadequately balancing exploration and exploitation phases. The suggested HGPSO effectively combines the advantageous features of Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) to overcome these limitations. The strong global exploration capabilities of GA, which successfully preserve variety and avert premature convergence, are integrated with the swift local exploitation and convergence attributes of PSO. A new multi-objective fitness function specifically designed for sensor deployment issues is created, facilitating the effective handling of trade-offs between conflicting objectives. The efficacy of the HGPSO approach is meticulously assessed in seven consistent situations and practical applications, encompassing environments with intricate impediments. A comparative examination is performed against six prominent metaheuristic algorithms acknowledged in literature. Results indicate that HGPSO regularly surpasses these competing methods across all assessment categories. Regarding average fitness values, HGPSO exceeds POHBA by 14 %, MAOA by 20 %, IDDT-GA by 21 %, EFSSA by 29 %, CFL-PSO by 35 %, and OBA by 45 %. These findings underscore HGPSO's exceptional theoretical framework and validate its practical relevance for extensive, real-world IoT implementations. By adeptly utilizing the exploration capabilities of GA and the exploitation strengths of PSO, HGPSO becomes a highly versatile and resilient optimization framework, making substantial contributions to addressing the deployment issues of next-generation IoT and WSN.