* 802.11e starts taking shape
At the Internet Telephony show in Los Angeles earlier this month, it was fitting that the subject of QoS came up often. During the Wi-Fi conference track, someone joked that the IEEE 802.11e QoS standard “has been six to nine months away for five to six years.”
At the Internet Telephony show in Los Angeles earlier this month, it was fitting that the subject of QoS came up often. During the Wi-Fi conference track, someone joked that the IEEE 802.11e QoS standard “has been six to nine months away for five to six years.”
Indeed, mid-2005 is now the expected time frame for full 802.11e approval.
However, last month, the Wi-Fi Alliance did begin taking action, similar to what it did when the sorely needed 802.11i security standard was a while in coming. It began certifying product interoperability among the completed components of the standard so that the industry could begin benefiting from some of the advancements.
In the case of QoS, it is the packet-prioritization portion – dubbed Wi-Fi Multimedia, or WMM, by the alliance – which is ready for prime time.
At this juncture, the only WLAN system to be WMM-certified is Cisco’s Aironet 1200 802.11a/g AP; other certified products are system components, such as Wi-Fi chips and software. However, many Wi-Fi vendors already support the long-proven Spectralink Voice Priority (SVP) scheme in their APs.
The availability of the WMM capability does open up the ability to prioritize traffic to makers of consumer-electronics devices such as DVD players and televisions, which might eventually be networked over Wi-Fi connections.
In addition, AP makers that have not implemented SVP could implement WMM for prioritization. For voice applications, their customers might still require Spectralink’s SVP server or NetLink Telephony Gateway in their configurations for admission control until the full 802.11e standard, which includes scheduled RF media access and possibly bandwidth provisioning, is implemented.
The role of SVP on the wired side is to coordinate the timing of voice calls when sending voice packets to and from an AP. Packets flow isochronously upstream and downstream to eliminate the possibility that traffic will collide.
WMM differs from SVP in that it works across different applications, not just voice. It is based on the IP Differentiated Services model without incurring IP’s overhead, and supports four priority levels.
The priority buckets default to voice as highest, then video, then best effort, then background. The background class, for example, might get print traffic and file downloads. You can, however, customize the settings to suit your own priorities.
