A Lightweight, Scalable and Distributed Admission Control Algorithm for Voice Traffic

P. Kulkarni, P. Dini, Sally McClean, Gerard Parr, Michaela Black

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

3 Citations (Scopus)


The idea of carrying voice traffic on IP networks has been found to be very lucrative. However to achieve a quality similar to that offered by the existing telephone networks, the IP network should be in a position to honour the stringent quality of service (QoS) requirements of voice traffic. If traffic in excess of the network capacity is admitted into the network, QoS may be violated resulting in performance degradation. One way of providing sustained and consistent QoS is by regulating the number of voice calls that are admitted into the network such that the load on the network is less than or equal to the capacity of the network. This mechanism is known as call admission control. This paper proposes a scalable and distributed call admission control algorithm that operates at the network edge and factors the local passive measurements into the admission decision. Performance evaluation of the proposed algorithm through ns2 simulations reveals that it is successful in detecting rate mismatches (input rate greater than output rate) and subsequently rejecting admission requests (as long as input rate is greater than output rate) thereby delivering on the QoS guarantees demanded by voice applications. Moreover, it is simple and lightweight from an implementation perspective.
Original languageEnglish
Title of host publicationUnknown Host Publication
Number of pages6
ISBN (Print)1-4244-0353-7
Publication statusPublished (in print/issue) - 2007
EventIEEE International Conference on Communications, 2007. ICC '07 - Glasgow, Scotland
Duration: 1 Jan 2007 → …


ConferenceIEEE International Conference on Communications, 2007. ICC '07
Period1/01/07 → …


Dive into the research topics of 'A Lightweight, Scalable and Distributed Admission Control Algorithm for Voice Traffic'. Together they form a unique fingerprint.

Cite this