A demand-assignment algorithm based on a Markov modulated chain prediction model for satellite bandwidth allocation

Francesco Delli Priscoli, Dario Pompili

Research output: Contribution to journalArticlepeer-review

4 Scopus citations

Abstract

This article deals with the problem of the design of a control-based demand-assignment algorithm for a satellite access network using a Markov modulated chain traffic prediction model. The objective is to guarantee a target Quality of Service (QoS) to Internet traffic, while efficiently exploiting the air interface. The proposed algorithm is in charge of dynamically partitioning the uplink bandwidth capacity in a satellite spotbeam among the in-progress connections. Such partition is performed aiming at matching the QoS requirements of each connection and maximizing the satellite bandwidth exploitation. A closed-loop Control Theory approach is adopted to efficiently tackle the problem of the delay between bandwidth requests and bandwidth assignments, while minimizing the signaling overhead caused by control messages. The algorithm efficiently copes with both the satellite propagation delay and the delays inherent in the periodic nature of the bandwidth request mechanism. The proposed demand-assignment algorithm and Markov chain traffic prediction model are shown to improve the overall satellite network performance through extensive simulation experiments.

Original languageEnglish (US)
Pages (from-to)999-1012
Number of pages14
JournalWireless Networks
Volume15
Issue number8
DOIs
StatePublished - Nov 2009

All Science Journal Classification (ASJC) codes

  • Information Systems
  • Computer Networks and Communications
  • Electrical and Electronic Engineering

Keywords

  • Control theory
  • IP traffic prediction
  • Internet access
  • Quality of service
  • Satellite networks

Fingerprint Dive into the research topics of 'A demand-assignment algorithm based on a Markov modulated chain prediction model for satellite bandwidth allocation'. Together they form a unique fingerprint.

Cite this