The co-operative network

April 2002

In the search for effective wireless broadband, mesh radio adopts a distinctive approach that provides a more efficient network as more customers sign up for services

The gnarled nut that the 'last mile' issue has become remains frustratingly difficult to crack. Network operators and service providers are increasingly beset by a wave of technologies that could potentially close the gap between their fibre trunk networks and a client base that is all too anxious for the industry to accelerate the rollout of broadband. While the established vendors of copper-based DSL and fibre-based cable are finding new business, many start-up operators, discouraged by the high cost of entry into wired markets, have been looking to evolving wireless radio and laser options.

One relatively late entrant into this competitive mire is mesh radio, a technology that has quietly emerged to become a potential holder of the title 'next big thing'. Although still at an early stage of operator tests, mesh has won sufficient attention that the analyst consultancy Strategis Group has forecast the market could reach Û11.4 billion as soon as 2003.

Based in Cambridge in the UK, Radiant Networks is seen as the leading light of the mesh radio movement. Founded only in 1997, it has patents pending for the systems behind its 'MeshWorks' solution. Trials are commencing in Frankfurt, Germany in partnership with Star 21; in Spain through Jazztel subsidiary Banda 26; and in Cardiff, Wales, with BT partnering (see the 'Mesh radio in Cardiff' box). US firm Caly Networks has developed an alternative mesh solution that differs from Radiant's in its practice. Its equipment is being assessed by Energis in Scotland.

The mesh concept

Mesh radio is entering a wireless market currently dominated by local multipoint distribution or LMDS solutions, which rely on the installation of base stations linked to the trunk network. Each customer in the same region accesses the same base station. The US has seen the most enthusiastic take-up of LMDS services to date, with cable proving too expensive to lay in remote areas.

The challenges LMDS systems must overcome include high initial capital expenditure and a coverage model less well suited to rural areas than to urban communities -- where operators inevitably face price competition from DSL and cable carriers subject to the same economic forces.

A mesh-based network rejects the idea of a sophisticated and costly base station in favour of a co-operative system. The concept is that each customer's wireless receiver can function as a relay point or node for other receivers within range, enabling a data signal to pass through several nodes as it progresses from the trunk network to its destination.

Viri Patel, VP for business strategy and marketing at Radiant Networks, says: 'The great characteristic of mesh is that rather than the operator having to build an infrastructure, each customer adds capacity to the infrastructure. As your number of links increases, your capacity increases.'

Under Radiant's system, each customer node installation has four antennae, each pointing in a different direction. A limited degree of movement enables each connector to line itself up with a connector in another node to achieve the best link.

Conceptually, then, mesh radio combines ideas from wide area networks and local wireless networks -- indeed, the US firm 'Mesh Networks' is developing unrelated technology that applies the mesh principle to the 802.11 cards increasingly deployed in laptops.

The beauty of the approach is that while LMDS networks become congested and face slowdown at peak times as more users join the system, mesh networks become more efficient as more customers in that region join. As the density in a region increases, resiliency improves, as nodes are more easily able to switch their signal along a different path if the closest neighbouring node is taken off-line.

The Radiant model allows for seed nodes, sites established by the operator to provide support for the first customers to join the network. Joining seed nodes in an operator's pre-launch preparation are Trunk Network Connection Points (TNCPs), the bridge between mesh and fibre systems, and Mesh Insertion Points (MIPs). These act both as subscriber nodes and as points of contact for TNCPs. The importance of MIPs is that they reduce the risk of bottlenecks occurring as networks become busier.

Customer installations use an Outdoor Unit or ODU, which relays data to an Indoor Unit or IDU, in turn the connection points for local networks or PCs. It is possible for several customers to share a single ODU to reduce installation costs for each.

The Caly Networks mesh system shares similar concepts of co-operative nodes, although the execution is different in many details. The job of the Radiant ODU's four antennae is performed by a device based on a Luneberg lens, a dome structure that delivers 360-degree coverage. Switching is achieved through IP rather than Radiant's use of the ATM protocol.

Made for high frequencies

Mesh radio's natural home looks to be in the 28 to 40GHz wavelength bands, which are currently the focus of radio authority licence tenders in Europe and the US. The frequencies have two key disadvantages for potential LMDS-based licensees: the signal reach falls as the frequency increases, and the signal becomes unable to pass through obstacles, meaning clear line of sight is essential.

Neither issue represents a drawback for mesh systems, Viri Patel claims. 'The industry believes reach will be around a kilometre, a kilometre and a half.' As the density of nodes in a region increases, the presence of alternative routers means that line of sight becomes less of a challenge. 'You can get to 95% coverage by having four nodes per square kilometre,' says Patel.

The assumption that distances between nodes will be short means that the maximum transfer speed of the MeshWorks solution is 25Mbps. Patel is careful to distinguish this figure from the real-world experience of customers, as their nodes handle traffic from neighbours as well as their own content.

'We can provide 600 customers per square kilometre with speeds of 4Mbps,' he says. 'That's the average... but it can burst to 25Mbps.' Once these parameters are established, however, a dense mesh network can cope readily with many simultaneous users without degrading transfer speeds, as is the case with DSL, cable and LMDS options.

Start-up costs

Because the connection point between a mesh radio network and the fibre trunk network is little different from any other node, initial capital costs for an operator can be significantly lower than under an LMDS scheme. The potential stumbling block to mesh radio's prospects, of course, is that each node is itself relatively sophisticated, which means customer start-up costs are high. 'Our nodes are around €8,000,' says Patel, 'and an indoor unit is typically about $1,000. Our target is to get to a consumer price of €1,100. We're confident that we'll be able to offer the right price point for consumers.'

The high start-up costs for customers in the near term, coupled with speeds double those of the fastest DSL implementation, mean that the initial market for mesh radio is seen as SMEs, with residential launches following in perhaps another 12 months. Operators are sure to encourage co-operative schemes, where firms or residents sharing a building have one node to deliver all the bandwidth they require.

The longer-term challenge for Radiant and other mesh equipment manufacturers is to ensure that common standards are agreed. How much more powerful might mesh become if nodes installed by different operators were able to relay information? 'We're quite keen to co-operate with other vendors to expand the standard,' Patel says.

Although trials are just starting, mesh radio appears a well-conceived technology, and the first commercial launches are anticipated for the end of the year or early 2003. It may be too far behind DSL and cable to have much impact in urban areas where these technologies are already established, but those regions currently suffering from under investment seem ripe for this solution.

top