Wireless meshes are an amazingly-flexible approach to addressing a wide variety of mobile – and fixed – network requirements. Here’s a quick introduction to the technology and applications of a truly innovative and unique technology.
What is a Wireless Mesh Network?
Wireless communications at its core is all about convenience – making it easy to apply a wide variety of network typologies quickly, easily, and cost-effectively. We often think of wireless in terms of a mobile device talking to a base station or access point – the point-to-multipoint topology, plus handoff between cells. And, normally, the interconnections between those cells use wire or perhaps a different form of point-to-point wireless. Suppose instead that the required interconnect was implemented as a form of relay, with one cell simply redirecting traffic wirelessly to the next, making it possible to implement almost any configuration; that’s the domain of the wireless mesh.
Think of it this way: imagine that the mobile device is talking to a base station or access point, but that base station or AP isn’t directly connected to backhaul, wired or wireless. Rather, traffic is forwarded, often over the same wireless technology as is used by a client, to another base station, and perhaps even over several hops. That, then, is a wireless mesh. Using this technique, it’s possible to deploy even large-scale networks for almost any application without having to worry about the wire (or, again, specialized wireless) otherwise required to interconnect the infrastructure. As long as a given station can be powered, and has the appropriate mesh software (typically, firmware) required, it can serve as a node in a mesh. Meshes thus have the key advantages of rapid, low-cost deployment, easy scalability, and the flexibility to handle essentially any application.
Let’s consider an example from the wireless-LAN world, where most enterprise-class vendors already offer mesh as an option. Suppose Wi-Fi coverage is required at a given location, but there’s no wire or cable to that spot. In other words, it’s not possible to plug the required AP into the wired LAN as long as power is still required (think solar cells and batteries as a last resort). All network administrators have to do is to put the AP into mesh mode (this is done via the management console), and voilà – user traffic relays through the AP. Coverage can be established quickly and easily, without a lot of advance planning, and meshes also have a high degree of fault-tolerance. In the event that a relay point is congested or suddenly becomes unavailable, another can quickly assist or even take over. And overall capacity is increased with each additional mesh node.
If some similarity is seen here between a wireless mesh and, for example, the basic architecture of the Internet, such isn’t coincidental. The Internet proved that mesh architectures, wireless or not, are in fact the right way to implement large-scale networks.
Downsides to Wireless Mesh Networks
There is a potential downside – it’s important to point out that throughput in a mesh might be impacted if traffic must proceed through a number of hops, as latency can be added with each relay. Configuring two radios in the AP “back to back”, though, minimizes latency. Many applications regardless involve relatively small amounts of data, and aren’t as impacted by additional hops. This is particularly true in sensor-based applications, like telemetry. With proper design and configuration, and sophisticated (and often proprietary) mesh protocols that adapt to changing load conditions, sufficient capacity is almost always available in properly-configured wireless meshes.
Not Just about Infrastructure
Meshes aren’t just about infrastructure, by the way – we might also use meshes to interconnect client devices as well. Purely peer-to-peer wireless links have major challenges as the number of nodes grows, but meshes improve with scale. Imagine, just for example, every car on the road equipped with a mesh node. Wireless traffic would thus relay from car to car until reaching its destination. More cars mean more capacity, so that morning traffic jam might provision gigabits of capacity to the folks in the back seat just trying to get a head start on their day. A given car might signal the dealership when it detects a fault, which might in fact be remedied by new firmware downloaded over that same wireless mesh – no need to stop by the service department at all! It’s even possible for handsets and similar devices to work in mesh mode, or as a hybrid client/infrastructure mesh. There’s an obvious question of one’s battery being used in the service of another user, but it can work to the benefit of everyone.
For now, a few experiments with a Wi-Fi-based wireless mesh should illustrate the power, flexibility, and performance of contemporary mesh architectures, and these are easy to do without disrupting a production WLAN. And, once these are demonstrated, production deployments usually aren’t far behind.
By Craig Mathias
Craig J. Mathias is a principal with Farpoint Group, a wireless and mobile advisory firm in Ashland, Mass. Founded in 1991, the company works with manufacturers, network operators, enterprises and the financial community on technology assessment and analysis, strategy development, product design and marketing, education, training and the integration of emerging technologies into business operations.
Craig, an internationally recognized expert on wireless communications and mobile computing technologies, is a well-known and often-quoted industry analyst and frequent speaker at industry conferences and events. He is a member of the advisory board for Interop, chairs the conference’s wireless and mobility track and also co-chairs the 2012 Mobile Connect conference, plus blogs and writes columns for several publications and websites. Craig holds an Sc.B. degree in applied mathematics/computer science from Brown University and is a member of the IEEE and the Society of Sigma Xi.