DocumentCode :
2581058
Title :
Connection-level performance analysis for adaptive bandwidth allocation in multimedia wireless cellular networks
Author :
Nasser, Nidal ; Hassanein, Hossam
Author_Institution :
Sch. of Comput., Queen´´s Univ., Kingston, Ont., Canada
fYear :
2004
fDate :
2004
Firstpage :
61
Lastpage :
68
Abstract :
In this paper, we propose an adaptive bandwidth framework for supporting multiple classes of multimedia services with different quality of service (QoS) requirements in the next generation of wireless cellular networks. The framework combines the following components: (i) a threshold-based bandwidth allocation policy. (ii) an efficient threshold-type call admission control (CAC) algorithm, and (iii) a bandwidth adaptation algorithm (BAA). The framework can be modeled as a multi-dimensional Markov chain, and therefore, a product form solution is provided. The QoS metrics - new call blocking probability, handoff call dropping probability, and degradation probability are derived. The analytical results are supported by simulation and show that this work improves the service quality by minimizing the handoff call dropping probability.
Keywords :
Markov processes; bandwidth allocation; cellular radio; multi-access systems; multimedia communication; probability; telecommunication congestion control; adaptive bandwidth allocation; bandwidth adaptation algorithm; call admission control; call blocking probability; degradation probability; handoff call dropping probability; multidimensional Markov chain; multimedia wireless cellular networks; quality of service; Bandwidth; Channel allocation; Degradation; Intelligent networks; Land mobile radio cellular systems; Next generation networking; Performance analysis; Quality of service; Telecommunication traffic; Wireless networks;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Performance, Computing, and Communications, 2004 IEEE International Conference on
Print_ISBN :
0-7803-8396-6
Type :
conf
DOI :
10.1109/PCCC.2004.1394944
Filename :
1394944
Link To Document :
بازگشت