Intel and Cisco gang up on mesh
When elephants dance...
Intel and Cisco are attempting to set their seal on a key emerging area of wireless technology, mesh networking, by pushing a new standard.
Intel’s keen interest in mesh is well documented and it has demonstrated applications using Wi-Fi and UltraWideBand meshes. But the actual products to date have come from a host of start-ups, which may now be threatened by the determination of Wi-Fi’s biggest players to ensure that the development of the market follows their agenda.
A study group for the proposed mesh standard will meet in January at the IEEE 802 meeting in Vancouver, Canada. The group is led by engineers from Intel and Cisco, though is open to all members of the 802 group, but the hands of the giants are clearly to be seen steering the process.
Intel claims a standard will prevent the fragmentation of the market between the various proprietary offerings of the start-ups such as Firetide, BelAir and Strix. It also believes the IEEE process will lead to a more efficient mesh approach for Wi-Fi. "The 802.11 protocol was not designed with mesh as a primary use, the MAC is inefficient across multiple hops, and my fear is most meshes would be inefficient and waste the spectrum available," said Intel’s lead engineer on the project, Steve Conner, who is developing a mesh protocol at Intel that would span 802.11 and UltraWideBand, and which could be the basis of a future standard.
As with WiMAX, an IEEE standard will help to kickstart the development and adoption of mesh technologies by guaranteeing interoperability and future proofing to customers, and by bringing down prices.
Mesh is a broad term covering technologies that enable ad hoc networking - any capable PC or wireless device can become a router/repeater for other devices within range, with signals ‘hopping’ between them on the most efficient route available. In fixed mesh, these hops eventually take the user to a fixed base station or access point; in mobile peer-to-peer mesh, PCs or handhelds can be equipped to act as hotspots or access points themselves.
The usual arguments rage, with specialists claiming that a standard will impose a lowest common denominator and that the process will just be the Intel/Cisco roadmap in disguise. Ike Nassi, founder of Firetide, said: “It's true we don't interoperate with other mesh networks, but there are a lot of things we do that the other guys don't to make mesh networking easy to use”.
Intel has certainly become adept at driving and using the standards process in the wireless market, preferring to ensure that it has critical influence on industry specifications than to create de facto alternatives. Cisco, for its part, has pushed several de facto standards and has sought to ensure that it is in the driving seat where choice of technology for standards is concerned, whether de facto or industry – witness the Cisco Compatible Extensions program.
We are likely to see a hybrid approach to mesh, with the two giants rolling out products and signing up partners to support their de facto platforms, thus making these almost certainly to form the core of the IEEE specification, which could take three years to complete.
The two companies represent an almost unchallengeable combination and both have an equally critical need to dominate the WLAN sector as they face the decline of their traditional markets, with the world shifting from wired PCs to mobile smart devices. Not only does a concerted effort put them several steps ahead of competitors, but they know that, with their strengths in the channel and the industry as a whole, they can accelerate the uptake of WLANs more rapidly than industry associations can.
The two companies already caused uproar earlier this year when they formed a partnership to integrate Intel Centrino and Cisco’s Linksys consumer devices more tightly, causing accusations of attempts to hijack Wi-Fi and exclude smaller players. Now they are widely seen to be doing the same in the nascent mesh market. Both moves fit in with Intel’s and Cisco’s historic approach to standards – wholehearted support if they can dominate the process, as Intel is doing with 802.16/WiMAX; if not, offering their own technologies and using their industry weight to make these de facto.
But there is no attempt to circumvent the IEEE process, and there is a real danger that the mesh start-ups, however clever their technology, will lack the unity and weight to make the platform widely appealing, and will fragment a promising approach with proprietary products.
The start-ups that have attracted the most attention in recent months, and which potentially have the most to lose from the Cisco-Intel tie-up, include PacketHop, Firetide and BelAir. There have also been announcements from companies such as Ember, which are focusing on short range wireless technologies such as UltraWideBand for their mesh frameworks. This is an area where Intel is also developing aggressively and the projected initiatives are unlikely to remain confined to Wi-Fi for long.
The start-ups have different approaches – BelAir is focusing on extending Wi-Fi’s outdoor range using mesh techniques in order to create hotzones and work more effectively in tall buildings; while Firetide, PacketHop and Strix are rolling out wireless routers and other devices for creating self-configuring, ad hoc networks without access points, mainly targeting the enterprise. Intel itself plans a wireless chipset with an integrated access point next year, which will lend itself to mesh networks in the business or the digital home.
For Intel, mesh is an approach that can make wireless networking easier to deploy and more ubiquitous and therefore can stimulate demand for all its key mobile technologies, and so it has placed a once obscure, hobbyist platform at the heart of its strategy. Its support will undoubtedly spur the uptake of mesh and the development of workable, affordable products, but it will also destroy the dream of the mesh pioneers – of a technology geared to community networks and, like Linux, communally owned - and put a new way of interacting firmly in the hands of the giants.
© 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