Development of Data and Computing Convergent Network
The data and computing convergent network (DCCN) is an intelligent communication network infrastructure tailored for data space applications. It promotes data space construction, data elements circulation, and integration of computing power and data. It also provides technical support for economic g...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | zho |
| Published: |
《中国工程科学》杂志社
2025-02-01
|
| Series: | 中国工程科学 |
| Subjects: | |
| Online Access: | https://www.engineering.org.cn/sscae/EN/PDF/10.15302/J-SSCAE-2024.10.005 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The data and computing convergent network (DCCN) is an intelligent communication network infrastructure tailored for data space applications. It promotes data space construction, data elements circulation, and integration of computing power and data. It also provides technical support for economic growth related to data rights, data circulation, and data transactions. This study defines the DCCN, outlines the key functions of its data plane, control plane, and orchestration layer, reviews the macro development requirements of the DCCN, and discusses the current development status and international trends of DCCN technologies. It further studies key DCCN components and technologies, including end sides, internal networking of data centers, data center exits, communication among data centers, networking among computing centers, networking between data and computing centers, control layers, orchestration layers, and security systems. Moreover, the study introduces application scenarios and cases of DCCNs, including east‒west data-center interconnections, computing power networks in urban areas, industrial extranet interconnections, energy facility interconnections, and industry-scale large models. Research also covers the challenges of developing DCCNs, based on which, we propose the following suggestions for developing the DCCN: (1) establishing public specialized networks that support the high-quality development of industry-scale large models; (2) promoting the construction of DCCN-based scientific facilities, serving national scientific development; (3) using the DCCN to promote applications in data spaces; (4) carrying out large-scale collaboration among computing facilities to overcome the bottleneck of insufficient computing capability at a single point. The suggestions aim to provide references for the development of DCCN infrastructures. |
|---|---|
| ISSN: | 1009-1742 |