A Survey of Compute First Networking

Gong, Xiaomin; Bai, Chenchen; Ren, Shuangyin; Wang, Jingchao; Wang, Chunjiang

doi:10.1109/icct59356.2023.10419572

Search citation statements

Order By: Relevance

Paper Sections

Select...

Citation Types

Supporting

Mentioning

Contrasting

Year Published

2024

Publication Types

Select...

Other2

Article1

Relationship

Self Cite0

Independent3

Authors

Journals

Cited by 3 publications

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

Graph-Classifier-Aided SnF Scheduling Method for Optical Computing Power Networks

Lin,

Xiao,

Zheng

et al. 2024

2024 IEEE/CIC International Conference on Communications in China (ICCC)

View full text Add to dashboard Cite

Graph-Classifier-Aided SnF Scheduling Method for Optical Computing Power Networks

Lin,

Xiao,

Zheng

et al. 2024

2024 IEEE/CIC International Conference on Communications in China (ICCC)

View full text Add to dashboard Cite

Task Offloading and Resource Allocation for ICVs in CFNs via Constrained MARL

Wang,

He,

Liu

et al. 2024

2024 IEEE International Symposium on Product Compliance Engineering - Asia (ISPCE-ASIA)

View full text Add to dashboard Cite

Dynamic Resource Aggregation Method Based on Statistical Capacity Distribution

Wang,

You,

2024

Electronics

View full text Add to dashboard Cite

The computing network is a novel architecture that enables resource matching through the network. In distributed computing networks, computing resource management devices collect resource information and report it to network nodes. These nodes then broadcast the information to guide resource matching. One challenge is efficiently aggregating and disseminating computing resource information, as directly reporting fully multi-dimensional data can cause excessive overhead, while overly simplified aggregation may reduce matching accuracy. Existing aggregation methods typically rely on static resource information, overlooking the heterogeneity and dynamics of computing resources that arise from variations in resource capabilities and fluctuations over time, leading to suboptimal matching decisions. In response, this study proposes a dynamic resource aggregation method based on statistical capacity distribution. By modeling the capacity distribution of computing nodes, this method captures dynamic resource information, enabling more precise resource matching. Additionally, constructing resource groups and calculating representative distributions effectively compress the volume of data announcements. Experiments and data analysis demonstrate that, compared to static resource matching methods, the proposed method improves matching accuracy by 48%. Furthermore, it reduces announcement overhead by approximately 77.1% compared to existing dynamic resource allocation methods. These findings provide an efficient solution for resource aggregation in distributed computing networks.

show abstract

A Survey of Compute First Networking

Cited by 3 publications

References 12 publications

Graph-Classifier-Aided SnF Scheduling Method for Optical Computing Power Networks

Graph-Classifier-Aided SnF Scheduling Method for Optical Computing Power Networks

Task Offloading and Resource Allocation for ICVs in CFNs via Constrained MARL

Dynamic Resource Aggregation Method Based on Statistical Capacity Distribution

Contact Info

Product

Resources

About