A Seamless Adaptive Video Streaming Technique Based on the Double DQN 


Vol. 45,  No. 10, pp. 1700-1707, Oct.  2020
10.7840/kics.2020.45.10.1700


PDF
  Abstract

With the recent spread of smartphones, HTTP Adaptive Streaming (HAS)-based video streaming services are becoming more common. The key to HAS-based streaming technology is to choose the video quality that is suitable for dynamic changing network environments and various terminal conditions, making it difficult to consider various environmental variables if only procedural algorithms are used. To compensate for this, various methods of introducing artificial intelligence techniques are being tried, and this paper proposes a Double DQN(DDQN)-based video quality selection algorithm. This method addresses the disadvantages of existing Deep Q-Network(DQN)-based research that although it effectively increases Quality of Experience(QoE), it takes a long time to create a learning model and over-estimation problems occur during the learning process. It also shows that learning model creation time can be reduced by about 21% while choosing more appropriate video quality.

  Statistics
Cumulative Counts from November, 2022
Multiple requests among the same browser session are counted as one view. If you mouse over a chart, the values of data points will be shown.


  Cite this article

[IEEE Style]

M. Choi, S. Jung, K. Lim, "A Seamless Adaptive Video Streaming Technique Based on the Double DQN," The Journal of Korean Institute of Communications and Information Sciences, vol. 45, no. 10, pp. 1700-1707, 2020. DOI: 10.7840/kics.2020.45.10.1700.

[ACM Style]

Minje Choi, Sungwook Jung, and Kyungshik Lim. 2020. A Seamless Adaptive Video Streaming Technique Based on the Double DQN. The Journal of Korean Institute of Communications and Information Sciences, 45, 10, (2020), 1700-1707. DOI: 10.7840/kics.2020.45.10.1700.

[KICS Style]

Minje Choi, Sungwook Jung, Kyungshik Lim, "A Seamless Adaptive Video Streaming Technique Based on the Double DQN," The Journal of Korean Institute of Communications and Information Sciences, vol. 45, no. 10, pp. 1700-1707, 10. 2020. (https://doi.org/10.7840/kics.2020.45.10.1700)