A New Block Fetching Scheme Considering Traffic Smoothing in Multimedia Servers

Sooyong Kang*, Heon Y. Yeom

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

1 Citation (Scopus)

Abstract

Typical block placement schemes generally assume CTL (Constant Time Length) block size, one block read for each stream in a round, round-robin striping, and peak rate-based admission control. Traditional smoothing schemes for stored continuous media objects do not consider the block layout of the storage system. Hence, the combination of schemes from the two domains can introduce new challenges. In this paper, we present a server transmission delay problem that arises when traditional block placement and smoothing schemes are used at the same time in a continuous media server. To resolve the problem, we first present two simple straightforward solutions, Server-side Block Prefetching and Multiple Block Read, and then propose a new solution Smoothed Fetching. SF overcomes the defects of the other two schemes by exploiting a smoothing technique when retrieving blocks from disks and using a tight admission control algorithm. Simulation results show that SF achieves the best performance among the three solutions.

Original languageEnglish
Pages (from-to)303-328
Number of pages26
JournalMultimedia Tools and Applications
Volume22
Issue number3
DOIs
Publication statusPublished - Mar 2004
Externally publishedYes

Keywords

  • Block fetching
  • Block placement
  • Multimedia system
  • Traffic smoothing

Fingerprint

Dive into the research topics of 'A New Block Fetching Scheme Considering Traffic Smoothing in Multimedia Servers'. Together they form a unique fingerprint.

Cite this