Service Interval Optimization with Delay Bound Guarantee for HCCA in IEEE 802.11e WLANs

Augusto FORONDA
Yuhi HIGUCHI
Chikara OHTA
Masahiko YOSHIMOTO
Yoji OKADA

Publication
IEICE TRANSACTIONS on Communications   Vol.E90-B    No.11    pp.3158-3169
Publication Date: 2007/11/01
Online ISSN: 1745-1345
DOI: 10.1093/ietcom/e90-b.11.3158
Print ISSN: 0916-8516
Type of Manuscript: PAPER
Category: Network
Keyword: 
wireless network,  bandwidth optimization,  bounded delay,  scheduler,  

Full Text: PDF(1MB)>>
Buy this Article


Summary: 
IEEE 802.11e Medium Access Control (MAC) is a supplement to the IEEE 802.11 Wireless Network (WLAN) standard to support Quality of Service (QoS). The 802.11e MAC defines a new coordination function, namely Hybrid Coordination Function (HCF), which takes the QoS requirements of flows into account and allocates Transmission Opportunity (TXOP) to stations. On the basis of mean sending rate, delay of Variable Bit Rate (VBR) traffic cannot be bounded with the reference HCF scheduling algorithm proposed in this supplement. In this paper, we propose a new Connection Admission Control (CAC) and a scheduling algorithm that utilize the token bucket and a modified Latency-Rate (LR) scheduling algorithm to guarantee a bounded delay for HCF Controlled Channel Access (HCCA). The new Service Interval (SI) is calculated to optimize the number of stations accommodated and takes into account delay bound and token bucket parameters. We show that it is possible to obtain worst-case performance guarantees on delay. First, we analyze the behavior of the new scheduler with a loss free wireless channel model and after this, with a burst loss model and we explain how it is possible to extend this scheduler for a multi-rate scheme. Properties of the proposal are investigated both theoretically and using ns-2 simulations. We present a set of simulations with both Constant Bit Rate (CBR) and VBR flows and performance comparisons with HCF scheduling algorithm. The results show that the delay upper bound can be achieved for a large range of networks load with bandwidth optimization.