방송공학회논문지 (Journal of Broadcast Engineering)

  • 제25권4호
  • /
  • Pages.609-619
  • /
  • 2020
  • /
  • 1226-7953(pISSN)
  • /
  • 2287-9137(eISSN)
한국방송∙미디어공학회 (The Korean Institute of Broadcast and Media Engineers)
권호 표지
DOI QR코드

DOI QR Code

영상 품질 및 전송효율 최적화를 위한 심층신경망 기반 영상전송기법

Video Transmission Technique based on Deep Neural Networks for Optimizing Image Quality and Transmission Efficiency

  • 투고 : 2020.05.20
  • 심사 : 2020.07.09
  • 발행 : 2020.07.30
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

고품질 비디오 스트리밍 요구에 따라 제한된 대역폭에서 높은 전송률이 필요하고, 트래픽 혼재 상황이 더 발생한다. 특히 실시간 영상 서비스를 제공 시 패킷 손실 및 비트 오류 확률이 더 크게 증가한다. 이러한 문제를 해결하기 위해 실시간 서비스 품질향상을 위한 방법으로 FEC 기술의 한 종류인 랩터 코드가 어플리케이션 영역에서 활발히 사용되고 있다. 본 논문에서는 랩터 코드를 활용하여 유사한 수준의 화질에서 전송 효율을 높이기 위한 다양한 심층 신경망(Deep Neural Network, DNN) 기반 영상전송 파라미터를 결정하는 방법을 제안한다. 제안된 신경망은 패킷 손실율(Packet Loss Rate), 비디오 인코딩 속도 및 전송속도를 입력으로 사용하고 랩터 FEC 파라미터와 패킷 크기를 출력으로 한다. 제안한 방법은 기존 멀티미디어 전송 기법과 유사한 수준의 PSNR(Peak Signal-to-Noise Ratio)에서 전송 효율을 최적화하여 평균 1.2% 높은 스루풋(throughput)을 보였다.

In accordance with a demand for high quality video streaming, it needs high data rate in limited bandwidth and more traffic congestion occurs. In particular, when providing real time video service, packet loss rate and bit error probability increase significantly. To solve these problems, a raptor code, which is one of FEC(Forward Error Correction) techniques, is pervasively used in the application layers as a method for improving real-time service quality. In this paper, we propose a method of determining image transmission parameters based on various deep neural networks to increase transmission efficiency at a similar level of image quality by using raptor codes. The proposed neural network uses the packet loss rate, video encoding rate and data rate as inputs, and outputs raptor FEC parameters and packet sizes. The results of the proposed method present that the throughput is 1.2% higher than that of the existing multimedia transmission technique by optimizing the transmission efficiency at a PSNR(Peak Signal-to-Noise Ratio) level similar to that of the existing technique.

키워드

참고문헌

  1. Cisco. White Paper, "Cisco VNI Forecast and Methodology," 2015-2020.
  2. SANTOS-GONZALEZ,Ivan, et al.Implementation and analysis of real-time streaming protocols. Sensors, 2017, 17.4: 846. https://doi.org/10.3390/s17040846
  3. Kwon, Oh Chan, Yunmin Go, and Hwangjun Song. "An Energy-Efficient Multimedia Streaming Transport Protocol Over Heterogeneous Wireless Networks." IEEE Trans. Vehicular Technology 65.8 (2016): 6518-6531. https://doi.org/10.1109/TVT.2015.2475281
  4. Motorola White Paper, "Realistic LTE Performance from Peak Rate to Subscriber Experience," 2009.
  5. M. Luby, "LT codes," in Proceedings of the Annual Symposium on Foundations of Computer Science, pp. 271-280, 2002.
  6. A. Shokrollahi, "Raptor codes," IEEE Tr. on Information Theory, Vol. 52, No. 6, pp. 2551-2567, June, 2006. https://doi.org/10.1109/TIT.2006.874390
  7. O. C. Kwon and H. J. Song, Raptor Codes-based Multipath Multimedia Transport Protocol over Wireless Networks, Proceedings of the Korea Information Science Society Conference, 1243-1245, 2014
  8. GO, Yunmin; SONG, Hwangjun. Raptor codes-based energy-efficient screen mirroring system. In: Visual Communications and Image Processing (VCIP), 2016. IEEE, 2016. p. 1-4.
  9. KWON, Oh Chan, et al. MPMTP: Multipath multimedia transport protocol using systematic Raptor codes over wireless networks. IEEE Transactions on Mobile Computing, 2015, 14.9: 1903-1916. https://doi.org/10.1109/TMC.2014.2364042
  10. Jong-Man Lee, (2018). A New Parameter Determination Method for Image Transmission protocol Using Deep Neural Network. Journal of the Institue of Electronics Engineers, 55(1), 30-36.
  11. Nair, Vinod, and Geoffrey E. Hinton. "Rectified linear units improve restricted boltzmann machines." Proceedings of the 27th international conference on machine learning (ICML-10). 2010.
  12. Kingma, Diederik, and Jimmy Ba. "Adam: A method for stochastic optimization.", Published as a conference paper at the 3rd International Conference for Learning Representations, San Diego, 2015
  13. Y. Bengio, A. Courville, and P. Vincent., "Representation Learning: A Review and New Perspectives," IEEE Trans. PAMI, special issue Learning Deep Architectures, 2013
  14. J. Schmidhuber, "Deep Learning in Neural Networks: An Overview", Neural Networks, Vol 61, pp 85-117, Jan 2015 https://doi.org/10.1016/j.neunet.2014.09.003
  15. I. J. Kim, Deep Learning: A new trend in machine learning. Journal of the Korean Institute of Communication Sciences (Information and Communication), 31 (11), 52-57, 2014
  16. Csaji, Balazs Csanad. "Approximation with artificial neural networks." Faculty of Sciences, Etvs Lornd University, Hungary 24.48 (2001)