Dapper: Deploying Service Function Chains in the Programmable Data Plane Via Deep Reinforcement Learning
IEEE Transactions on Services Computing · 01.07.2023 · 16 (4) · 2532-2544
DOI: 10.1109/tsc.2023.3237244 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Services Computing, erschienen am 01.07.2023. Verfasst von Xiaoquan Zhang und 4 weiteren Autor:innen. Die Arbeit wurde bisher 16-mal zitiert (Crossref) und nennt 32 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Science and Technology Program of Guangzhou, National Key Research and Development Program of China.
Abstract (Originalsprache)
Network functions perform specific packet processing on network traffic. To meet operators’ needs, forming service function chains (SFCs) is a fundamental technique used in today's ISPs and datacenter networks. Implementing SFCs in the programmable data plane with high throughput and low latency is a new approach to satisfy demands of ever-growing network traffic. Previous works have proposed different solutions to solve the problem, but they all inevitably have to make trade-offs between running time and performance. For example, an ILP (Integer Linear Programming) can optimize cost but suffers from long running time in large-scale network topologies. Heuristic algorithms depend strongly on manual designs and usually have a performance gap with the optimal solution. In this paper, we propose Dapper , a framework for deploying SFCs in the programmable data plane using DRL (Deep Reinforcement Learning) with graph convolutional network. In order to expand the searching space to prevent the optimal value from being missed, Dapper allows the RL (Reinforcement Learning) agent to simultaneously extract features from both the substrate network and the hardware pipeline, and exploit a graph convolutional network to enhance performance. Moreover, a mask mechanism is also designed to accelerate Dapper and improve its scalability. Dapper has been implemented and extensively evaluated on both P4 hardware switches (equipped with Intel Tofino ASIC) and software switches (i.e., bmv2). Experimental results show that Dapper can automatically generate deployment solutions in a few seconds of running time after training. They also demonstrate that Dapper reduces hardware stage usage and the latency of SFCs by up to 17.8% and 50 $\sim$ ∼ 73% respectively on average when compared with heuristics.
Förderung
- National Natural Science Foundation of China
- Science and Technology Program of Guangzhou
- National Key Research and Development Program of China
- Engineering and Physical Sciences Research Council
- InnovateUK
- Guangdong Key Lab of AI and Multi-modal Data Processing
- United International College
- Guangdong Key Laboratory of Data Security and Privacy Preserving
Zitieren
APA
Zhang, X., Cui, L., Tso, F. P., Li, Z., & Jia, W. (2023). Dapper: Deploying Service Function Chains in the Programmable Data Plane Via Deep Reinforcement Learning. IEEE Transactions on Services Computing, 16(4), 2532-2544. https://doi.org/10.1109/tsc.2023.3237244
Vancouver
Zhang X, Cui L, Tso FP, Li Z, Jia W. Dapper: Deploying Service Function Chains in the Programmable Data Plane Via Deep Reinforcement Learning. IEEE Transactions on Services Computing. 2023;16(4):2532-2544. doi:10.1109/tsc.2023.3237244
BibTeX
@article{Zhang2023,
author = {Zhang, Xiaoquan and Cui, Lin and Tso, Fung Po and Li, Zhetao and Jia, Weijia},
title = {Dapper: Deploying Service Function Chains in the Programmable Data Plane Via Deep Reinforcement Learning},
journal = {IEEE Transactions on Services Computing},
year = {2023},
volume = {16},
number = {4},
pages = {2532--2544},
doi = {10.1109/tsc.2023.3237244}
}
EUVIMED