Solving Problems with Wireless Mesh Networks

Mesh networks can solve this problem and a bunch of other problems, aswell. The underlying technology borrows heavily from lessons learnedthrough the implementation of dynamic routing protocols on the Internet.MeshNetworks wins the prize for best name in this industry sector, but anumber of other companies, including BelAir, Firetide, Strix and Tropos,are busy selling interesting solutions. The mesh movement got a boostlast week when the IEEE created a study group to begin exploring a meshstandard. Along with the companies mentioned above, the development drewinterest from industry heavyweights Cisco and Intel. It's anybody'sguess how long it might take to develop a standard, but three to fiveyears seems like a safe estimate.

Mesh technology has both indoor and outdoor applications. Insidebuildings, products from Firetide and Strix can be used to create WLANsthat don't require Ethernet backhaul. In the simplest ad hocimplementation, you just plug a bunch of mesh-capable APs intoelectrical outlets in your building and the APs automatically formbackhaul connections amongst themselves. Of course, simple is not alwaysoptimal, so some products provide the ability to manage the mesh moreclosely, selecting appropriate links and radio technologies.

Outside the building LAN, meshes can be used to provide wireless accessacross wide geographic areas measured in miles rather than meters.Tropos has garnered some attention recently because of work it is doingwith NASA. Based on tests in the Arizona desert, NASA was able to deploywireless connectivity covering a two-square mile area using two to threeTropos devices. In urban areas, where buildings cause RF reflections,you might need 10 times as many devices. But for some applications,including public safety and Internet access in suburban or rural areas,the technology has significant potential. BelAir also offers outdoormesh products, which include multiple radios per access device, and theyaren't limited exclusively to 802.11.

The major benefits of wireless meshes include system resiliency,flexible coverage areas and rapid deployment. It's likely that evenmainstream WLAN infrastructure vendors will begin to employ some limitedmesh capabilities in their products to provide more deploymentflexibility in locations where cables are hard to pull (warehouses,historic buildings, etc.).

The challenges also are notable and include cost, security,interoperability and scalability. Vendors will need to educateprospective customers on all these issues. For example, while devicecost is likely to be higher than traditional WLAN systems, the true costof deployment needs to factor in both savings on cable installationcosts and user benefits associated with rapid deployment. Scalabilityand security issues are likely to be manageable, but they will add tooverall system complexity. Interoperability will need to wait forstandards, but as long as the systems are compatible with 802.11 at theaccess layer, many IT professionals may be willing to accept proprietarymesh implementations.

While most of today's products use the mesh to provide backhaul betweenAPs and into the Internet or intranet, in the future, the meshesthemselves are likely to extend out to client devices. In fact, a numberof universities are actively involved in designing wireless gridnetworks, which allow heterogeneous client devices to dynamically joinwireless networks, acting as both client devices and routers for otherdevices. Wireless grids may not be practical for many years, but meshnetworks are here today.

-- Dave Molta, [email protected]