2003UTS ePRESS (University of Technology Sydney)Requires access

Class-based fair intelligent admission control over an enhanced differentiated service network

Ming Li, Doan B. Hoang, Andrew Simmonds

Open publisher page 0 citations

Abstract

Integrated Service (IntServ) can provide powerful QoS on a per flow basis but it requires routers to perform per flow admission control and maintain per flow state. Differentiated Service (DiffServ) model is more scalable, but cannot provide service that is comparable to IntServ. To achieve scalability and a strong service model, we believe that DiffServ equipped with scalable admission control is a possible solution. This paper examines some fair intelligent admission control schemes over the enhanced DiffServ [1]. Specifically, we propose two admission control schemes: random early dropping admission control and random early remarking admission control. We apply these admission control schemes to address traditional DiffServ problem concerning fairness. The simulation results demonstrate our schemes can offer a class-based fair resource sharing. Such admission control schemes can also help enforce the desired service assurances. © Springer-Verlag Berlin Heidelberg 2003.

About this research paper

What this paper is about

Integrated Service (IntServ) can provide powerful QoS on a per flow basis but it requires routers to perform per flow admission control and maintain per flow state. Differentiated Service (DiffServ) model is more scalable, but cannot provide service that is comparable to IntServ. To achieve scalability and a strong service model, we believe that DiffServ equipped with scalable admission control is a possible solution. This paper examines some fair intelligent admission control schemes over the enhanced DiffServ [1]. Specifically, we propose two admission control schemes: random early dropping admission control and random early remarking admission control. We apply these admission control schemes to address traditional DiffServ problem concerning fairness. The simulation results demonstrate our schemes can offer a class-based fair resource sharing. Such admission control schemes can also help enforce the desired service assurances. © Springer-Verlag Berlin Heidelberg 2003.

Why it matters

A significance statement is not available in the OpenAlex record.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Integrated Service (IntServ) can provide powerful QoS on a per flow basis but it requires routers to perform per flow admission control and maintain per flow state. Differentiated Service (DiffServ) model is more scalable, but cannot provide service that is comparable to IntServ. To achieve scalability and a strong service model, we believe that DiffServ equipped with scalable admission control is a possible solution. This paper examines some fair intelligent admission control schemes over the enhanced DiffServ [1]. Specifically, we propose two admission control schemes: random early dropping admission control and random early remarking admission control. We apply these admission control schemes to address traditional DiffServ problem concerning fairness. The simulation results demonstrate our schemes can offer a class-based fair resource sharing. Such admission control schemes can also help enforce the desired service assurances. © Springer-Verlag Berlin Heidelberg 2003.

Key concepts: Admission control, Computer science, Quality of service, Scalability, Computer network, Differentiated services, Integrated services, Service (business)

Related papers

Back to paper searchBrowse research topicsOriginal source
Class-based fair intelligent admission control over an enhanced differentiated service network — Research Paper | ScholarLens