Wireless meshes with its future
- 12 May, 2003 22:00
It's time we had a new buzzword. Thanks to wireless technology proponents, that new poster child has arrived: mesh networking. You could easily write off the technology as a logical extension of wireless and IP-based routing technologies. But that would be a limited view of its potential.
The idea of mesh networking started gaining momentum in 2002, but now that Intel Corp. and Microsoft Corp. have joined the discussion initiated by startups such as MeshNetworks Inc., the technology promises to give users access to a rich set of applications. In many respects, mesh networks promise to deliver enterprises the wireless infrastructure they've always craved.
Existing wireless deployments using 802.11 freed us from the tangled mess of blue wire, engineering cable guys out of their jobs. But under the mesh networking model, a variety of clients, including handhelds and notebooks, would become both wireless consumers of data and wireless access points. Every device can be a router, which fosters a dynamic multihop environment that pundits claim will provide users with more robust wireless networks: Network traffic will no longer be dependent on a system with a single point of failure. This creates room for extending the "always-on" Internetto collaborative applications, including IM and e-mail,running on anything from handhelds to consumer devices. The network effect of this grid is also promising. I imagine it will not take long for wireless vendors to start posturing that it can help break down that dreaded last-mile telecommunications connection. The visionaries will argue that if a mesh networks system can bridge the spaces between home and the office, it can rival cell phone and fixed line networks.
But before the anti-carrier argument takes off, the isolated test cases need to prove themselves.
According to an Intel presentation at January's IDF show, early adopters will be home users. Consumer environments are ideal because of the close proximity of different computing and electronics devices. This example still relies on some type of last-mile high-speed Internet connection, of course. But what's most telling about Intel's perspective is not the adoption by consumers, but the closed environment in which these grids will thrive.
Wireless, like cell-phone technology, has distance limitations. So a short hop between clients inside the home means stronger and faster connections. The same idea applies equally and just as powerfully to the corporate campus.
The introduction of wireless mesh networks in the enterprise will have an interesting impact on collaboration and peer-to-peer communications. Microsoft has not detailed much of its vision in this space, but we do know the company's researchers are investigating how client software can leverage this style of communication.
You can understand Microsoft's interest given the distributed nature of software such as Office and the forthcoming InfoPath, and its investment in p-to-p desktop collaboration outfit Groove Networks Inc.
But Microsoft does not need to deliver solutions tomorrow. The hardware side of mesh networking is still in its infancy, with many technical hurdles still to be overcome. This includes maintaining quality of service, avoiding interference between wireless channels, security, and authentication. In addition, standards such as 802.11 continue to evolve, and other technologies such as Intel's AODV (Ad Hoc On-Demand Distance Vector) routing compete with a proposal from the Internet Engineering Task Force called DSR (Dynamic Source Routing).
Development challenges notwithstanding, mesh networking will emerge to become an increasingly vital part of the wireless evolution.
Sometimes there's more to buzz words than just the buzz. -- InfoWorld (US)