Toward Sustainable Edge Computing Infrastructures: A VEC-MEC-Cloud Collaborative Framework for Adaptive Resource Distribution

Edge computing systems experience significant challenges in dynamically adapting to unpredictable user mobility and fluctuating service demands, particularly in smart city scenarios where latency-sensitive applications are predominant. This paper addresses this critical gap by proposing a collaborative framework integrating vehicular edge computing (VEC), mobile edge computing (MEC), and cloud computing to achieve adaptive edge resource distribution. The system comprises core components including a network latency model, resource allocation model, deployment location identification model, and dispatch scheduling model. We evaluate our approach using real-world city maps and base station deployment information, demonstrating the effectiveness and robustness of the proposed methods. Results show an average error rate of approximately 0.02% in meeting time-sensitive applications' resource requirements and delay constraints, and a conservative 8.1% improvement in service availability compared to static deployment strategies, validating the framework's adaptability in dynamic environments. By enabling adaptive resource distribution, this approach offers a sustainable and cost-efficient solution for edge computing infrastructure, addressing the evolving demands of urban digital ecosystems.