Dynamic Edge Caching Strategy Integrating Deep Learning and Swarm Intelligence

Aiming at the problem that cache strategies are difficult to adapt to changeable factors in complex and dynamic edge caching environments, this study proposes a novel cache decision-making framework that deeply integrates deep learning prediction, user preference mining, and hybrid intelligent optimization. By constructing a user request prediction model and combining it with an improved collaborative filtering recommendation mechanism, a personalized cache candidate set is generated. Based on the prediction results, a hybrid optimization strategy that integrates the discrete particle swarm optimization algorithm and the genetic algorithm is designed to effectively avoid the trap of local optimality. Finally, a dynamic adaptive update mechanism is established. By monitoring the change rate of content popularity through a sliding window, the real-time evolution of the cache strategy is achieved. Simulation experiments show that the proposed method significantly outperforms traditional strategies in key indicators such as cache hit rate and user request latency, verifying the effectiveness of the algorithm in dynamic and complex scenarios.