Article Preview
TopIntroduction
In the last decades, we observe an increasing interest in wireless technologies that offer mobility, high data rates and high availability. Wireless networks based on the IEEE 802.11 standard (IEEE Standards Association, 2007) have spread over all kinds of markets, because of their lower deploying and maintaining costs, easiness of deployment and great flexibility to users.
Recently, service and equipment providers and universities have studied and deployed a kind of wireless network, called wireless mesh network (Bruno et al., 2005). Wireless mesh networks are built over a multi-hop forwarding wireless backhaul (Aoun et al., 2006) that, opposed to the IEEE 802.11 ad hoc networks, does not have restrictions regarding mobility, because the backbone nodes are usually placed in a fixed location.
One of the approaches to deploy wireless mesh networks is using the network layer ad hoc routing protocols, like the ones deployed in the Fluminense Federal University (UFF) (Campista et al., 2008), Massachusetts Institute of Technology (MIT) (Bicket et al., 2005), and in the so called digital cities, like Seattle (USA) (http://www.seattlewireless.net/) and Dublin (Ireland) (Weber et al., 2003).
Nowadays we can observe media convergence as a global tendency. Therefore, those networks have to carry different kinds of traffic with different quality of service (QoS) requirements. Applying QoS in the Internet is already a challenge because of its distributed nature and the best effort mechanism of its base protocol, IP (Internet Protocol). The wireless medium brings even more challenges, considering its limited bandwidth, error-prone radio channels, shadow areas, and signal and climate interferences. So, providing QoS for wireless networks has become a research area of great interest, resulting in standards like IEEE 802.11e (IEEE Standards Association, 2007).
Many commercial products already implement some mechanisms of the standard or a subset of them launched by WiFi Alliance, called WiFi Multimedia (WMM) (Wi-Fi Alliance, 2004). However, studies in the literature (Ramos et al., 2005) state that the values suggested in the standard for medium access parameters are not suitable for all kinds of wireless networks.
The standard does not map access parameters to network characteristics, such as traffic pattern, network load, mobility pattern, and so on. The responsibility of applying the QoS policy on the network, calculating and announcing the parameters, and adjusting them to the current status of the network, lies over the wireless access point (AP). Traffic classification and prioritization in ad hoc networks are mentioned, but they are out of the scope of the standard.
This work investigates the influence of each IEEE 802.11e access parameter (AIFS, CWmin, CWmax, and TXOPLimit) in QoS metrics, such as throughput, end-to-end delay, collision rate and loss rate, over wireless mesh networks. This work also proposes two new configuration settings for those access parameters: a static one and another that includes a mechanism for dynamic adjustment of the TXOPLimit parameter. The main goal is to provide voice traffic service guarantee in wireless mesh networks, keeping its end-to-end delay below 150ms and its loss rate between 1% and 3% (Kawata & Yamada, 2007), without degrading significantly the performance of the other traffic classes. This work is an extended version of the paper of Gerk and Muchaluat-Saade (2012) presented on the 2012 International Conference on Communications and Information Technology (ICCIT).
The remainder of this paper is structured as follows. The next section overviews key aspects of IEEE 802.11e standard mechanisms. The following one presents related work found in the literature. The fourth section explains in detail the proposed configuration settings. The fifth section evaluates the proposal through simulation experiments, comparing them with the standard. Finally, the last section concludes the paper suggesting future work.