A refined Greylag Goose optimization method for effective IoT service allocation in edge computing systems

The growth of the Internet of Things (IoT) has intensified the need for efficient service placement in edge computing environments. This problem remains challenging due to dynamic workloads and heterogeneous resources. Existing swarm intelligence algorithms, such as QPSO-SP and WOA-FSP, often struggle to balance exploration and exploitation effectively. We propose the Modified Greylag Goose Optimization (MGGO) algorithm, which introduces adaptive mechanisms for dynamic population partitioning, stagnation detection, and learning-based control. MGGO optimizes key performance metrics, including energy consumption, latency, throughput, and load balancing. Experimental evaluations on synthetic service placement workloads show that MGGO achieves 12-15% improvement over GGO, QPSO-SP, BOA, and WOA-FSP across all metrics. These findings demonstrate MGGO's potential for improving edge service placement in dynamic environments.