Analysis of a markov-modulated fluid model for multimedia traffic with loss and delay priorities

Marwan Krunz, Hennan Hughes

Research output: Contribution to journalArticlepeer-review

3 Scopus citations

Abstract

This paper presents a complete fluid-flow analysis of a traffic control mechanism in an ATM multiplexer, known as Nested Threshold Cell Discarding with Multiple Buffers. NTCDIMB is a buffer management scheme that implements loss and delay priority queueing strategies in order to satisfy a wide range of quality-of-service requirements. We consider a heterogeneous mix of traffic which consists of real-time traffic with a stringent cell delay requirement, and nonreal-time traffic with a stringent cell loss requirement. Moreover, a proportion of real-time cells has an additional stringent loss requirement. Cell arrivals are mode led using a fluid-flow modeling approach. The system we analyze consists of two buffers interacting via a common server. The steady-state probability distribution for the buffers' content as well as other performance statistics are obtained. Our approach is based on computing the conditional probability distributions of the buffers and using this information to find the desired unconditional stationary distributions. This approach can be used, in general, to analyze multiple dependent systems of buffers. Numerical examples based on the analysis are used along with simulations to study the performance of NTCDIMB. It is shown that NTCDIMB can be used effectively to provide performance guarantees on different cell loss and delay requirements of the traffic.

Original languageEnglish (US)
Pages (from-to)309-329
Number of pages21
JournalJournal of High Speed Networks
Volume3
Issue number3
DOIs
StatePublished - Jan 1 1994
Externally publishedYes

Keywords

  • Buffer Management
  • Fluid Analysis
  • Multiplexing
  • Priorities

ASJC Scopus subject areas

  • Information Systems
  • Hardware and Architecture
  • Computer Networks and Communications

Fingerprint Dive into the research topics of 'Analysis of a markov-modulated fluid model for multimedia traffic with loss and delay priorities'. Together they form a unique fingerprint.

Cite this