Characterization of Movie Contents and Its Impact for Traffic Design

• KOIKE Arata

  Traffic Engineering Division, NTT Advanced Technology Corp.

• TAKIGAWA Satoko

  Traffic Engineering Division, NTT Advanced Technology Corp.

• TAKEDA Kiyoka

  Traffic Engineering Division, NTT Advanced Technology Corp.

• KOBAYASHI Akihisa

  NTT Cyber Space Laboratories, NTT Corporation

• MORIMOTO Masashi

  NTT Cyber Space Laboratories, NTT Corporation

• KAWASHIMA Konosuke

  Traffic Research Center, NTT Advanced Technology Corp.

Abstract

In this paper, we first investigate the characteristics of movie contents over the Internet. As in the previous studies, we found the lognormal-distribution well fits the distribution of file size for the whole set of general movie contents. When we specifically focus on the subset that consists of movie trailers, however, it shows different distribution from the lognormal-distribution. Our analysis shows it is similar to an exponential-distribution. We here assume that movie trailers are one of the relevant contents for Contents Delivery Networks (CDN) or Peer-to-Peer (P2P) file exchange communities. We further studies the relationship between playing duration and file size for the movie trailers and we did not find any linear correlation among them. We next consider bandwidth requirements to retrieve movie trailer contents. Our objective is to make it possible for user to view the contents in real-time. Many previous studies investigate bandwidth requirement based only on the file size distribution. In this paper, we analyze the traffic design criteria for CDN or P2P by taking into account both of the results for the file size distribution and the relationship between playing duration and file size for movie trailers. Simulation studies reveal the impact for the bandwidth requirement.

Journal

• IEICE Trans. on Communications

  IEICE Trans. on Communications 86 (6), 1839-1848, 2003-06-01

  The Institute of Electronics, Information and Communication Engineers

Keywords

• movie trailer
• file size
• playing duration
• peer-to-peer
• simulation
• traffic design

Details

• CRID: 1570009752411473664
• NII Article ID: 110003221885
• NII Book ID: AA10826261
• ISSN: 09168516
• Text Lang: en
• Data Source:
  • CiNii Articles