Hybrid GA-PSO Optimization for Controller Placement in Large-Scale Smart City IoT Networks

The Internet of Things (IoT) plays an important role in the development of smart cities. IoT forms a large network, and optimal controller placement plays a crucial role in ensuring network performance and resilience. This paper proposes a hybrid optimization approach that combines Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) to strategically place controllers. Kaunas (Lithuania) was selected as a real-world smart city model. A large-scale Narrowband Internet of Things (NB-IoT) network with 2000 nodes was simulated, and 10 controllers were optimally placed in the network to minimize latency, balance load, enhance energy efficiency, and redundancy. The performance of the proposed hybrid GA-PSO algorithm was compared with random and K-Means clustering placements under three scenarios: normal operation, node failures, and traffic spikes. Simulation results demonstrate that the hybrid approach outperforms the other two methods in terms of load balancing, packet loss, energy efficiency, scalability, and redundancy. These findings highlight the robustness and effectiveness of the proposed hybrid algorithm in optimizing controller placement for smart city environments.