Intelligent optimization of wireless sensor networks based on hybrid prediction-driven multi-sink elastic deployment, QoS-constrained routing decision, and energy-latency dual-objective optimization

In order to reduce the contradiction between node failure and energy consumption and delay in wireless sensor networks, and ensure the timeliness and reliability of data transmission, this paper designs an optimization framework. The framework integrates and enhances three optimization algorithms: Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Non-dominated Sorting Genetic Algorithm II (NSGA-II). Specifically, the GA is enhanced by adopting a “region-based crossover” strategy and initializing the population based on Sink node density, which improves network fault tolerance and stability. The PSO is improved by applying an adaptive mechanism to dynamically adjust learning factors and adding a QoS constraints to ensure routing decisions meet quality of service constraints. The NSGA-II is enhanced by using an adaptive selection mechanism and an objective-oriented crossover strategy, which balance energy consumption and latency optimization. Through simulation experiments, the optimized framework performs better than the Federated double deep Q-network (Federated DDQN) and the Adaptive multi-objective particle swarm optimization (Adaptive MOPSO) in terms of energy consumption, delay, throughput and reliability in wireless sensor networks. It also performs well in both inverted generational distance (IGD) (0.03) and hypervolume (HV) (0.87), showing advantages in the dual-objective optimization of energy consumption and latency. The research in this paper is of great significance for improving the scalability, stability, and efficiency of wireless sensor networks in practical applications. It provides the theoretical basis and practical guidance for intelligent optimization and adaptive scheduling in more complex network environments in the future.