Multimedia 802.11e standard speeds up
Taking power back from vendors
The Wi-Fi Alliance says it will skip an interim certification for multimedia and voice over IP extensions to 802.11x, since it seems that the IEEE standards body will fast track the full standard, 802.11e, to be approved by June.
The body, which tests and promotes WLAN standards, had been planning to release a stopgap spec – in a two-step process similar to that for Wi-Fi security extensions, by which the Alliance released WPA as a move towards the full IEEE 802.11i standard.
The interim standard for multimedia would have been Wireless Multimedia Extensions but it will be unnecessary if 802.11e can be released by the IEEE as early as June. Frank Hanzlik, managing director of the Wi-Fi Alliance, said that once 802.11e is ratified, “the alliance will begin offering a certification for 802.11e and all of its components."
WME supports quality of service features such as priority tagging and queueing and is a subset of Wi-Fi Scheduled Multimedia, which will be incorporated in ‘e’. But the full standard is badly needed as interest in VoIP heats up, with compression issues particularly important to resolve for the standard – especially as WME was to have been frozen, not evolved into 802.11e, probably causing upgrade complexities.
The original media access control (MAC) protocol for Wi-Fi does not support differentiation of different traffic types or sources, making it unsuitable for applications where certain traffic needs to be prioritized – such as voice or video over IP. In 802.11 in its current state, there are two modes of communications, both of which will be enhanced in 802.11e. With Distributed Coordination Function (DCF), based on ‘listen before talk’ technology, a wireless station waits for a quiet period on the network before transmitting data and detecting any collisions.
An optional second mode, Point Coordination Function (PCF), goes a step further as it supports time sensitive traffic. It splits the time into contention-free and contention periods and transmits data during the former. However, while these two modes offer coordination and time sensitivity, neither distinguishes between different types of traffic.
The proposed 802.11e standard would add extensions to both modes to support quality of service for voice applications, while retaining backwards compatibility with existing Wi-Fi variants. DCF would be enhanced with support for eight different traffic categories, with lower priority categories of traffic waiting for the others to go first, before accessing the medium.
However, there are no guarantees of service, which could still limit the viability of heavy duty voice over IP implementations.
Another enhancement would be to PCF, introducing a Hybrid Coordination Function that polls stations during contention-free periods and grants each station a specific start time and maximum duration for transmission. Since PCF has not been widely used, this second enhancement has received lower interest levels than Enhanced DCF, although the two can work together.
Earlier this year it appeared that these ‘e’ extensions would be delayed, possibly into next autumn. As with the security standard, a group of vendors, under the auspices of the Wi-Fi Alliance, came together to propose an interim specification, Wireless Multimedia Enhancements (WME). This was ostensibly to ensure that end users who were eager to implement voice and multimedia applications over Wi-FI would have at least a measure of reassurance that their early deployments would be compatible with future standards.
Like WPA in security, WME was basically a snapshot of the current state of the full standard. Like 802.11e, it would synchronise the client device and access points in terms of how they handle queueing, channel access and collision avoidance, so there is no need to configure the devices specially to recognise VoIP phones and their traffic. WME uses only four priority level rather than the eight of 802.11e.
However, there will be relief in many quarters if WME becomes unnecessary. Such moves by the vendors arouse suspicion for many reasons and are becoming a worrying trend. With the Wi-Fi market becoming increasingly crowded and the pressure on margins rising, suppliers need ways to persuade the enterprise sector, in particular, to revive its flagging interest and investment in WLAN – much of the downturn caused itself by confusion over standards.
The danger is that they rush equipment to market based on half-baked specifications, which will still require upgrades and a second round of spending by the users once the full standard emerges. In doing so, they not only increase the scepticism of enterprises about wireless deployments, but they also start to rob the IEEE of its standards role.
Interim standards are even more susceptible to the influence of one or two powerful vendors than the full standards process; the Wi-Fi Alliance is an industry body not a standards organization; and once a stopgap technology has been adopted, this puts huge pressure on the IEEE to make it the basis of the new standard, even if this is not the best technical solution.
It is no surprise that, as with WPA, the chief mover behind WME is Microsoft, closely followed by Cisco – the two companies most criticised in the Wi-Fi world for seeking to subvert the standards process by pushing their own technologies. The Wi-Fi Alliance is open about favoring the large suppliers, claiming it needs to base its testbed on big companies because
“we know they'll be around for a while”.
All this is nothing new of course. Standards bodies such as the IEEE, ISO, ETSI and ITU have always specified pure specifications and then placed the actual commercial marketing, deployment and testing in the hands of consortia, and these have always been dominated by the companies with the time, money and loud voices to spare to become top tier members.
Consortia need money and they need the support of the big names in order to have teeth and to be credible with users. But the Wi-Fi Alliance is being more blatant than most in effectively excluding the small innovators, which could bring so much its processes.
Analysts at Meta Group recently slammed Cisco and Microsoft for pushing proprietary implementations, especially in the area of wireless security, rather than working to accelerate true standards. This, argued senior research analyst Bjarne Munch, was slowing the progress of IEEE specifications and delaying enterprise roll-outs rather than accelerating them, in turn is contributing to declining WLAN revenues in the enterprise space.
In the case of 802.11e, it does look as though the interim standard will be made redundant by the unaccustomed speediness of the progress of the full specification. There is a growing awareness among standards bodies that, to retain their role and relevance, they will have to move at a pace that is appropriate to the industry and user base they serve. The IETF recently adopted moves to speed up its particularly slow processes – under which a standard can take 10 years to complete – and the IEEE has also been attempting to streamline its work.
There are dangers to this – compromising quality to speed; and the opportunity for large suppliers to hijack a process where speed is of the essence. The recent fiasco surrounding the search for the IEEE 802.15.3a standard, based on UltraWideBand, shows how a change of approach aimed at streamlining can fall victim to supplier politics. Speeding up is not easy – this is a complex process. The reason for WME, ofr instance, was that the 802.11e working group reached something of a roadblock, partly because too many interests were represented. As a standards body, the IEEE needed to cover all possible interests in QoS, but these can be conflicting – QoS requirements for wireless high definition television makers, for instance, are very different to those for wireless Lan makers. While these issues were discussed, the delays gave the Wi-Fi Alliance the excuse to fast forward on WME. So there are risks to trying to crank up the standards process to match the speed of the industry.
But the alternative is for standards bodies to be increasingly sidelined as vendors get desperate to push out new products to a rapidly changing market, and so take matters into their own hands. The prospect that 802.11e will go ahead without requiring the interim specification gives some hope that the vendors cannot hijack standards too easily.
© Copyright 2003 Wireless Watch
Wireless Watch is published by Rethink Research, a London-based IT publishing and consulting firm. This weekly newsletter delivers in-depth analysis and market research of mobile and wireless for business. Subscription details are here.
Sponsored: IBM FlashSystem V9000 product guide