Optimizing Resource Utilization in Edge Computing Environment with Dynamic Load Balancing Scheduling Algorithm

Edge cloud computing extends cloud resources to the network edge, catering to the increasing demands of edge devices and applications. Traditional container orchestration engines, while suitable for edge environments, often rely on static scheduling algorithms that consider only CPU and Memory, inadequate for the diverse needs of edge applications. This paper proposes an enhanced scheduling algorithm that incorporates two new metrics: pod-rate and ephemeral-storage, offering a more comprehensive approach to resource optimization. Additionally, it introduces a dynamic weighting mechanism using the Dynamic Load Balancing Scheduling Algorithm, which adjusts resource weights in real time based on usage during Pod deployment. Experiments demonstrate that this algorithm reduces the standard deviation of comprehensive loads by 57% compared to standard scheduling algorithms under high cluster loads. This boost improves edge computing's load balance and resource efficiency.