Task Offloading Method for Edge-Cloud Collaboration in Resource-Constrained Environments

The processing of large-scale, compute-intensive tasks in vehicular networks poses significant challenges to edge computing resources. This paper proposes a task offloading method based on an improved NSGA-II algorithm for edge-cloud collaboration, establishing a dual-objective optimization model for latency and energy consumption to achieve dynamic allocation of computational tasks among onboard, edge, and cloud resources. This method first analyzes the latency tolerance characteristics of tasks and constructs a two-layer offloading decision mechanism. Secondly, it improves the core operations of the NSGA-II algorithm to optimize population evolution strategies. Finally, a resource allocation scheme based on task characteristics is designed. Experimental results demonstrate that this method effectively reduces the total consumption of the system, significantly enhances task response performance, and maintains stable operation under high-load scenarios. The research findings provide a new solution for optimizing task allocation in vehicular network environments and are of significant importance for improving the performance of edge computing systems.