Intelligence-Endogenous Management Platform for Computing and Network Convergence

Zicong Hong,Xiaoyu Qiu,Jian Lin,Wuhui Chen,Yue Yu,Hui Wang,Song Guo,Wen Gao

2023-08-07

Abstract:Massive emerging applications are driving demand for the ubiquitous deployment of computing power today. This trend not only spurs the recent popularity of the \emph{Computing and Network Convergence} (CNC), but also introduces an urgent need for the intelligentization of a management platform to coordinate changing resources and tasks in the CNC. Therefore, in this article, we present the concept of an intelligence-endogenous management platform for CNCs called \emph{CNC brain} based on artificial intelligence technologies. It aims at efficiently and automatically matching the supply and demand with high heterogeneity in a CNC via four key building blocks, i.e., perception, scheduling, adaptation, and governance, throughout the CNC's life cycle. Their functionalities, goals, and challenges are presented. To examine the effectiveness of the proposed concept and framework, we also implement a prototype for the CNC brain based on a deep reinforcement learning technology. Also, it is evaluated on a CNC testbed that integrates two open-source and popular frameworks (OpenFaas and Kubernetes) and a real-world business dataset provided by Microsoft Azure. The evaluation results prove the proposed method's effectiveness in terms of resource utilization and performance. Finally, we highlight the future research directions of the CNC brain.

Artificial Intelligence

What problem does this paper attempt to address?

### Problems the Paper Aims to Solve

The paper aims to address the following issues:

1. **Mismatch between Computing Resources and Demand**:
- Emerging applications (such as the metaverse, AI chatbots, and autonomous driving) are rapidly increasing the demand for computing power, but computing nodes are unevenly distributed geographically. Economically active regions have high computing demands, while resource-rich but sparsely populated areas have lower costs.
- The hardware heterogeneity of different computing nodes and the diversity of goals of emerging applications lead to an imbalance between supply and demand, thereby affecting service quality.

2. **Resource Coordination in Computing and Network Convergence (CNC)**:
- An intelligent management platform is needed to coordinate distributed resources and tasks in CNC, enabling users to easily access the required computing resources.
- The management platform needs to have four functional modules: perception, scheduling, adaptation, and governance, to achieve efficient and automated supply-demand matching.

3. **Prototype Implementation of Intelligent Resource Management**:
- A prototype system of the CNC brain was implemented based on deep reinforcement learning technology and evaluated by integrating the OpenFaas and Kubernetes frameworks.
- Experimental results show that this method performs excellently in terms of resource utilization and service quality.

In summary, the paper mainly focuses on how to use artificial intelligence technology to solve the problems of resource allocation and task scheduling in a computing and network convergence environment.