Skip to content

☕ Experimental/research project to build high-performance and energy-efficient virtual network functions (VNFs).

License

Notifications You must be signed in to change notification settings

stevelorenz/build-vnf

Repository files navigation

MIT Licensed

Build-VNF

Build-VNF

This is just an experimental, research and study oriented project to explore new ideas for my PhD thesis (in short, for papers ... 💡) .

About

Build-VNF (Build Virtualized Network Function) is a collection (monorepo 🙈) of sources and tools that I developed and used during my PhD study 😉 (2018-2022) at ComNets TU Dresden.

When I started my PhD in 2018, SDN/NFV was a totally new topic for me, so this monolithic repository was created to store and manage my prototype/experimental code during my studies. Since I learned that there are already many popular and mature research and industry projects for NFVI and NFV-MANO components, I decided to investigate the possibility of building fast, energy-efficient and intelligent VNFs for 5G and beyond 😆.

Subdirectories are currently created for sub-projects (ideas and topics) and measurements. Each subdirectory should contain clear READMEs for their usage.

The structure and code here is completely unstable right now, as I'm not an expert yet and am still in the try-error-retry loop 😢. I plan to create a separate repository after this one has some good and meaningful 🙈 results. Better, structured documentation will be added later when the project structure is stable.

Thanks to all the open source SDN/NFV projects for teaching me.

Project Structure

  • FFPP (Fast Forward Packet Processing): A fancy 🐼 packet processing engine (library) designed for CNF (Containerized Network Function) and COIN (Computing in the Network) use cases
  • COIN_DL: Implementing a COIN-driven distributed deep learning system for in-network video traffic compression
  • pktgen: Scripts and tools for software packet generators
  • scripts: Utility scripts and helpers

Citing Our Works

If you like any ideas or tools inside this repository, please consider reading and citing our papers. These papers have detailed description and rigorous performance evaluation results.

@inproceedings{xiang2018latency,
  title={Latency measurement of service function chaining on OpenStack platform},
  author={Xiang, Zuo and Gabriel, Frank and Nguyen, Giang T and Fitzek, Frank HP},
  booktitle={2018 IEEE 43rd Conference on Local Computer Networks (LCN)},
  pages={473--476},
  year={2018},
  organization={IEEE}
}

@article{xiang2019reducing,
  title={Reducing latency in virtual machines: Enabling tactile internet for human-machine co-working},
  author={Xiang, Zuo and Gabriel, Frank and Urbano, Elena and Nguyen, Giang T and Reisslein, Martin and Fitzek, Frank HP},
  journal={IEEE Journal on Selected Areas in Communications},
  volume={37},
  number={5},
  pages={1098--1116},
  year={2019},
  publisher={IEEE}
}

@article{xiang2021xman,
  author={Xiang, Zuo and Höweler, Malte and You, Dongho and Reisslein, Martin and Fitzek, Frank H.P.},
  journal={IEEE Transactions on Network and Service Management},
  title={X-MAN: A Non-intrusive Power Manager for Energy-adaptive Cloud-native Network Functions},
  year={2021},
  volume={},
  number={},
  pages={1-1},
  doi={10.1109/TNSM.2021.3126822}
}

@article{Wu2021njica,
  title = {Computing Meets Network: COIN-aware Offloading for Data-intensive Blind Source Separation},
  author = {Huanzhuo {Wu} and Zuo {Xiang} and Giang T. {Nguyen} and Yunbin {Shen} and Frank H. P. {Fitzek}},
  year = {2021},
  date = {2021-09-01},
  journal = {IEEE Network Magazine, Special Issue },
  keywords = {},
  pubstate = {published},
  tppubtype = {article}
}
  • For COIN_DL and ComNetsEmu:
@article{xiang2020comnetsemu,
  title = {An Open Source Testbed for Virtualized Communication Networks},
  author = {Zuo {Xiang} and Sreekrishna {Pandi} and Juan A. {Cabrera Guerrero} and Fabrizio {Granelli} and Patrick {Seeling} and Frank H. P. {Fitzek}},
  year = {2021},
  date = {2021-01-01},
  journal = {IEEE Communications Magazine},
  pages = {1--8},
  note = {(Accepted, 2020)},
  keywords = {},
  pubstate = {published},
  tppubtype = {article}
}

@article{xiang2021you,
  title={You Only Look Once, But Compute Twice: Service Function Chaining for Low-Latency Object Detection in Softwarized Networks},
  author={Xiang, Zuo and Seeling, Patrick and Fitzek, Frank HP},
  journal={Applied Sciences},
  volume={11},
  number={5},
  pages={2177},
  year={2021},
  publisher={Multidisciplinary Digital Publishing Institute}
}

Contributing

This project exists thanks to all people (all my students 😉) who contribute. The list of all known contributors.

Contact

License

This project is licensed under the MIT license.

About

☕ Experimental/research project to build high-performance and energy-efficient virtual network functions (VNFs).

Topics

Resources

License

Stars

Watchers

Forks