The Journal of Bigdata (한국빅데이터학회지)

Korea Bigdata Society (한국빅데이터학회)
권호 표지
DOI QR코드

DOI QR Code

Distributed AI Learning-based Proof-of-Work Consensus Algorithm

분산 인공지능 학습 기반 작업증명 합의알고리즘

  • 채원부 (한국항공대학교 컴퓨터공학과) ;
  • 박종서 (한국항공대학교 소프트웨어학과)
  • Received : 2022.05.02
  • Accepted : 2022.06.18
  • Published : 2022.06.30
Download PDF
⟨ Previous Next ⟩

Abstract

The proof-of-work consensus algorithm used by most blockchains is causing a massive waste of computing resources in the form of mining. A useful proof-of-work consensus algorithm has been studied to reduce the waste of computing resources in proof-of-work, but there are still resource waste and mining centralization problems when creating blocks. In this paper, the problem of resource waste in block generation was solved by replacing the relatively inefficient computation process for block generation with distributed artificial intelligence model learning. In addition, by providing fair rewards to nodes participating in the learning process, nodes with weak computing power were motivated to participate, and performance similar to the existing centralized AI learning method was maintained. To show the validity of the proposed methodology, we implemented a blockchain network capable of distributed AI learning and experimented with reward distribution through resource verification, and compared the results of the existing centralized learning method and the blockchain distributed AI learning method. In addition, as a future study, the thesis was concluded by suggesting problems and development directions that may occur when expanding the blockchain main network and artificial intelligence model.

대부분의 블록체인이 사용하는 작업증명 합의 알고리즘은 채굴이라는 형태로 대규모의 컴퓨팅리소스 낭비를 초래하고 있다. 작업증명의 컴퓨팅리소스 낭비를 줄이기 위해 유용한 작업증명 합의 알고리즘이 연구 되었으나 여전히 블록 생성 시 리소스 낭비와 채굴의 중앙화 문제가 존재한다. 본 논문에서는 블록생성을 위한 상대적으로 비효율적인 연산 과정을 분산 인공지능 모델 학습으로 대체하여 블록생성 시 리소스 낭비문제를 해결하였다. 또한 학습 과정에 참여한 노드들에게 공평한 보상을 제공함으로써 컴퓨팅파워가 약한 노드의 참여 동기를 부여했고, 기존 중앙 집중 인공지능 학습 방식에 근사한 성능은 유지하였다. 제안된 방법론의 타당성을 보이기 위해 분산 인공지능 학습이 가능한 블록체인 네트워크를 구현하여 리소스 검증을 통한 보상 분배를 실험 하였고, 기존 중앙 학습 방식과 블록체인 분산 인공지능 학습 방식의 결과를 비교하였다. 또한 향후 연구로 블록체인 메인넷과 인공지능 모델 확장 시 발생 할 수 있는 문제점과 개발 방향성을 제시함으로서 논문을 마무리 하였다.

Keywords

References

  1. 유영상. "[기술정책 이슈] 데이터 경제 시대의 새로운 공정 경쟁 이슈", 한국전자통신연구원, 2020.
  2. Nakamoto, Satoshi. "Bitcoin: A peer-to-peer electronic cash system." Decentralized Business Review, 2008.
  3. 유성민. "4 차 산업혁명과 블록체인: 데이터 경제중심으로." 한국통신학회지 (정보와통신) 37.2 pp.23-30.2020.
  4. Dwork, Cynthia, and Moni Naor. "Pricing via processing or combatting junk mail." Annual international cryptology conference. Springer, Berlin, Heidelberg, pp. 139-147, 1992.
  5. https://ccaf.io/cbeci/index/comparisons
  6. https://www.fnnews.com/news/202202241509203078
  7. Ball, Marshall, et al. "Proofs of work from worst-case assumptions." Annual International Cryptology Conference. Springer, Cham, pp. 789-819, 2018.
  8. Oyinloye, Damilare Peter, et al. "Blockchain Consensus: An Overview of Alternative Protocols." Symmetry 13.8, 2021.
  9. Vassilevska Williams, Virginia. "Hardness of easy problems: Basing hardness on popular conjectures such as the strong exponential time hypothesis (invited talk)." 10th International Symposium on Parameterized and Exact Computation (IPEC 2015). Schloss Dagstuhl-Leibniz-Zentrum fuer Informatik, 2015.
  10. Baldominos, Alejandro, and Yago Saez. "Coin. AI: A proof-of-useful-work scheme for block-chain-based distributed deep learning." Entropy 21.8, 2019.
  11. https://www.deepbrainchain.org/assets/pdf/DeepBrainChainWhitepaper_ko.pdf
  12. Bravo-Marquez, Felipe, Steve Reeves, and Martin Ugarte. "Proof-of-learning: a blockchain consensus mechanism based on machine learning competitions." 2019 IEEE International Conference on Decentralized Applications and Infrastructures (DAPPCON). IEEE, pp. 119-124, 2019.
  13. Lihu, Andrei, et al. "A proof of useful work for artificial intelligence on the blockchain." arXiv preprint arXiv:2001.09244, 2020.
  14. Subhlok, Jaspal, et al. "Exploiting task and data parallelism on a multicomputer." Proceedings of the fourth ACM SIGPLAN symposium on Principles and practice of parallel programming, pp. 13-22, 1993.
  15. 안신영, et al. "딥러닝 분산처리 기술동향." [ETRI] 전자통신동향분석 31.3, 2016.
  16. Mirhoseini, Azalia, et al. "Device placement optimization with reinforcement learning." International Conference on Machine Learning. PMLR, pp. 2430-2439, 2017.
  17. Kim, Jin Kyu, et al. "Strads: A distributed framework for scheduled model parallel machine learning." Proceedings of the Eleventh European Conference on Computer Systems, pp. 1-16, 2016.
  18. Wang, Wei, et al. "SINGA: Putting deep learning in the hands of multimedia users." Proceedings of the 23rd ACM international conference on Multimedia, pp. 25-34, 2015.
  19. Larochelle, Hugo, et al. "Exploring strategies for training deep neural networks." Journal of machine learning research 10.1, 2009.
  20. Gardner, Matt W., and S. R. Dorling. "Artificial neural networks (the multilayer perceptron)-a review of applications in the atmospheric sciences." Atmospheric environment 32.14-15, 1998.
  21. Krizhevsky, Alex, Ilya Sutskever, and Geoffrey E. Hinton. "Imagenet classification with deep convolutional neural networks." Advances in neural information processing systems 25, 2012.
  22. Mikolov, Tomas, and Geoffrey Zweig. "Context dependent recurrent neural network language model." 2012 IEEE Spoken Language Technology Workshop (SLT). IEEE, pp. 234-239, 2012.
  23. Abadi, Martin, et al. "{TensorFlow}: A System for {Large-Scale} Machine Learning." 12th USENIX symposium on operating systems design and implementation (OSDI 16), pp. 265-283, 2016.
  24. Park, Y. M., et al. "Deep Learning Model Parallelism." Electronics and Telecommunications Trends 33.4, 2018.
  25. Donovan, Alan AA, and Brian W. Kernighan. The Go programming language. Addison-Wesley Professional, 2015.
  26. Micali, Silvio, Michael Rabin, and Salil Vadhan. "Verifiable random functions." 40th annual symposium on foundations of computer science (cat. No. 99CB37039). IEEE, pp. 120-130, 1999.
  27. Szydlo, Michael. "Merkle tree traversal in log space and time." International Conference on the Theory and Applications of Cryptographic Techniques. Springer, Berlin, Heidelberg, pp. 541-554, 2004.
  28. Ripeanu, Matei. "Peer-to-peer architecture case study: Gnutella network." Proceedings first international conference on peer-to-peer computing. IEEE, pp. 99-100, 2001.
  29. Masse, Mark. REST API design rulebook: designing consistent RESTful web service interfaces. "O'Reilly Media, Inc.", 2011.
  30. Paszke, Adam, et al. "Pytorch: An imperative style, high-performance deep learning library." Advances in neural information processing systems 32, 2019.
  31. Deng, Li. "The mnist database of handwritten digit images for machine learning research [best of the web]." IEEE signal processing magazine 29.6, 2012.
  32. Coron, Jean-Sebastien, et al. "Merkle-Damgard revisited: How to construct a hash function." Annual International Cryptology Conference. Springer, Berlin, Heidelberg, pp. 430-448, 2005.
  33. Goodfellow, Ian, et al. "Generative adversarial nets." Advances in neural information processing systems 27, 2014.