Title

Call Admission Control And Scheduling Policies For Umts Traffic For Qos Provisioning

Abstract

The future of wireless data services seems very promising with the advent of third generation (3G) wireless carriers providing a wide variety of data services in conjunction with the existing voice services. To deal with the varied nature of these services, the Universal Mobile Telecommunication System (UMTS)has defined four QoS classes of traffic. However, there exists no service level agreement between the users and the providers on the quality of data services because of the subjectiveness associated with user satisfaction. In this paper, we model user irritation factors using Sigmoid functions which reflect users' sensitivity and tolerance to QoS provided. We propose call admission control (CAC) algorithms for the four QoS classes and a scheduling mechanism that are specifically designed for the UMTS traffic classes. The CAC and scheduling algorithms jointly attempt to provide a bounded delay for non-real time content. The proposed mechanism takes into consideration the UMTS classes, instantaneous channel conditions, short-term and long-term irritation factors of individual users, fairness within every class and throughput maximization. Simulation experiments conducted with synthetically generated traffic for all four classes demonstrate that the proposed schemes are able to maintain the QoS requirements while maximizing throughput and keeping the churn rate under control. © Springer-Verlag Berlin Heidelberg 2004.

Publication Date

1-1-2004

Publication Title

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Volume

3079

Number of Pages

566-578

Document Type

Article

Personal Identifier

scopus

DOI Link

https://doi.org/10.1007/978-3-540-25969-5_51

Socpus ID

35048861326 (Scopus)

Source API URL

https://api.elsevier.com/content/abstract/scopus_id/35048861326

This document is currently not available here.

Share

COinS