Research on Heterogeneous Resources Collaborative Scheduling Model in Edge Computing Environment

Edge computing, as a distributed computing model, has shown significant advantages in reducing latency and improving real-time performance. However, due to the centralized resource pool characteristics, traditional cloud computing scheduling models are difficult to adapt to the decentralized and heterogeneous edge environment, network fluctuations, task bursts and other characteristics, and have limitations such as lack of network awareness, insufficient heterogeneous adaptation, and static priority mechanism. To address the above issues, a Dynamic Adaptive Resource Scheduling Model (DARM) is proposed, which constructs a three-level hierarchical architecture to achieve a balance between global collaboration and local self optimization. This model innovatively introduces a multidimensional resource perception algorithm to obtain real-time node status through a three-layer monitoring system consisting of physical layer, network layer, and energy layer; Design a dynamic priority adjustment strategy to break through the limitations of traditional EDF algorithms with a single time dimension, and construct a three-dimensional evaluation model of time value density energy sustainability; Establish a task migration decision tree, achieve adaptive migration decision-making through dynamic threshold determination and multi factor correlation analysis, and avoid performance fluctuations caused by ineffective migration. Experiments have shown that by dynamically adjusting task priorities through a three-dimensional evaluation model, DARM significantly reduces task completion time, improves resource utilization, and optimizes migration decisions using decision trees to reduce ineffective migration in edge cloud heterogeneous environments.