EUVIMEDEUVIMED entdecken
Crossref

Accelerating Hardware/Software Combined Traffic Processing With Fast and Efficient Asynchronous Flow Offloading

Xijin Yin, Yuanwei Lu, Xin Zhang, Xingtong Lin, Shengli Zheng, Bangwen Deng, Xianneng Zou, Yachen Wang, Guo Chen

IEEE Transactions on Networking · 2026 · 34 · 4233-4248

DOI: 10.1109/ton.2026.3671366 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Networking, 2026. Verfasst von Xijin Yin und 8 weiteren Autor:innen. Die Arbeit nennt 56 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Major Special Project of Changsha Science and Technology Plan, National Key Research and Development Program of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Hardware/software (hw/sw) combined systems are necessary to meet modern clouds’ requirements for processing huge amounts of network traffic by efficiently offloading large flows to hardware. However, existing hw/sw flow offloading systems typically perform traffic statistics collection and large flow selection within a time window—a time-window-based approach. Their offloading decision of large flows is synchronized in the unit of a time window, which is mismatched to the asynchronous and dynamic nature of each flow’s sending rate. Additionally, the flow measurement and selection for large flows are decoupled in these solutions, leading to memory and CPU inefficiency. In this paper, we introduce TAO, a novel solution to the hw/sw combined flow offloading problem by asynchronously selecting and offloading flows based on flow table entries. TAO can react faster to the rapid dynamics of flows by taking actions at each table entry and is more efficient by coupling flow measurement and selection into the entry. We have implemented a full-fledged TAO prototype based on the P4 switch and DPDK. Testbed results demonstrate that TAO can offload ~16% more traffic to hardware, outperforming existing solutions by achieving $42 \times $ lower memory overhead. Meanwhile, it reduces software CPU utilization by 66.7% and cuts tail forwarding latency by 95.59% compared to state-of-the-art methods.

Förderung

  • National Natural Science Foundation of China
  • Major Special Project of Changsha Science and Technology Plan
  • National Key Research and Development Program of China
  • Major Special Project of Shenzhen Science and Technology Plan

Zitieren

APA

Yin, X., Lu, Y., Zhang, X., Lin, X., Zheng, S., Deng, B., Zou, X., Wang, Y., & Chen, G. (2026). Accelerating Hardware/Software Combined Traffic Processing With Fast and Efficient Asynchronous Flow Offloading. IEEE Transactions on Networking, 34, 4233-4248. https://doi.org/10.1109/ton.2026.3671366

Vancouver

Yin X, Lu Y, Zhang X, Lin X, Zheng S, Deng B, et al. Accelerating Hardware/Software Combined Traffic Processing With Fast and Efficient Asynchronous Flow Offloading. IEEE Transactions on Networking. 2026;34:4233-4248. doi:10.1109/ton.2026.3671366

BibTeX
@article{Yin2026,
  author = {Yin, Xijin and Lu, Yuanwei and Zhang, Xin and Lin, Xingtong and Zheng, Shengli and Deng, Bangwen and Zou, Xianneng and Wang, Yachen and Chen, Guo},
  title = {Accelerating Hardware/Software Combined Traffic Processing With Fast and Efficient Asynchronous Flow Offloading},
  journal = {IEEE Transactions on Networking},
  year = {2026},
  volume = {34},
  pages = {4233--4248},
  doi = {10.1109/ton.2026.3671366}
}
Kostenlos starten →