Application-Aware Resource Allocation Based on Benefit–Cost Ratio in Computing Power Network with Heterogeneous Computing Resources

• DOI: 10.3390/photonics10111273
• Publication Date: 2023-11-17T14:10:38Z
• Authors: Yahui Wang; Yajie Li; Jiaxing Guo; Yingbo Fan; Ling Chen; Boxin Zhang; Wei Wang; Yongli Zhao; Jie Zhang
• Keywords: Provisioning; Benchmark (surveying); Symmetric multiprocessor system; Blocking (statistics); Resource Management

The computing power network (CPN) is expected to realize the efficient provisioning of heterogeneous through collaboration between cloud and edge computing. Heterogeneous resources consist CPU, GPU, other types power. Different applications may have diverse requirements for resources, such as general applications, CPU-intensive GPU-intensive applications. Service providers are concerned about how dynamically provide different in a cost-effective manner, deploy more much possible with limited resources. In this paper, concept benefit–cost ratio (BCR) proposed quantify usage efficiency CPU GPU CPNs. An application-aware resource allocation (AARA) algorithm designed processing With massive simulations, we compare performance AARA benchmark. terms blocking probability, utilization, BCR, achieves better than simulation results indicate that tasks can be accommodated by reducing 3.7% probability BCR-based allocation.