Your browser is not supported. Please update it.

4 June 2007

Clearmesh looks for buyer, casting new doubts over FSO

Free space optics (FSO) technology is periodically hyped up as a wireless alternative to fiber, but after every burst of interest in the laser-based system, collapse follows. The latest example is the failure of a pioneer that had focused on using FSO in the municipal market, Clearmesh Networks. The company is looking for a buyer after the incubator that funded it, Idealab, pulled support – raising questions again over whether FSO will ever have a real role to play in wireless networks.

FSO has remained a niche technology since it was originally developed by space agency NASA in the 1970s. It uses line of sight lasers or LEDs to transmit voice, data and video at up to 2.5Gbps over distances of 4-5 kilometers in 60GHz frequencies, carrying any protocol. Through use of wavelength division multiplexing, supporters say speeds will increase to 10Gbps in the next 1-2 years, rivalling physical and wireless fiber rates. There was a flurry of interest in 2003-4, driven partly by a high profile project by AT&T to combine FSO with microwave (still a major potential way for the light-based system to gain greater market appeal). However, AT&T drew back, with the first hybrid products appearing instead from specialist Lightpointe, and the clutch of small suppliers has largely failed in its dream of targeting the broadband access market (NTT of Japan is one of the few majors that has used FSO for last mile applications). The access market is already lost to RF technologies such as WiMAX and Wi-Fi, but the FSO community remains hopeful that the boom in wireless data usage will create new opportunities for the technology in the backhaul.

Clearmesh’s troubles are a blow to this view, since it had targeted a boom sector, municipal wireless, aiming to combine FSO-based links with Wi-Fi or other access options (see Wireless Watch July 3 2006 edition). Clearmesh emerged on the back of US interest in ‘wireless fiber’, particularly when the FCC opened up the so-called millimeter wave bands, in 71-76GHz and 81-86GHz, for license-exempt use. While GigaBeam and Loea led the way in pioneering radio-based products for these frequencies, California start-up ClearMesh looked to combine WLan switching with wireless optics to offer cost efficient, easily deployed wireless backhaul with gigabit-plus capacity. By focusing on mesh rather than point-to-point patterns, it said it used multiple hops to route signals through the network, getting round the line of sight requirements of higher frequencies.

The first product from the company, the ClearMesh Metro Grid, was targeted at service providers that wanted to offer business class WISP services, particularly to small and medium enterprises, with rates up to 100Mbps. ClearMesh claimed to be unique in combining standards-based Lan switching with wireless optical transmission. ‘The ClearMesh 300 Node is a switching platform,’ explained the company. ‘It has an Ethernet switch with 2Gbps Ethernet capacity. Four of the Ethernet ports are copper and they are connected to optical transceivers.’ Its transceivers differed from traditional FSO by using LEDs, which have a wider beam than lasers, enabling them to be installed on relatively unstable units like light poles. RF back-up is used in case of fog, the main obstacle to FSO reliability.

Now incubator Idealab has decided it is no longer ‘in its strategic interest’ to continue support for Clearmesh. The start-up’s senior vice president of marketing Fima Vaisman admitted to reporters that he was uncertain how long the company can last on its current resources, though it is still supporting customers. Likely buyers would be firms specialized in metrozone mesh or emergency response systems and the company could attract a low-price purchase from a player in this sector such as Tropos, or even Motorola or Cisco.

Given the doubts over the viability of FSO, and the overcrowding of start-ups in the metrozone market, Clearmesh may prove unlucky, but it does apparently have real customers and trials, and recently enhanced its range by unveiling a new system specifically for the booming wireless video survelliance sector. And the company has some fairly convincing figures when comparing its meshed system with fiber, although these are unlikely to improve quickly unless it can achieve volume. Considering installation costs, the company uses $5,000 per node in its ROI calculations. In contrast to trenching fiber, it says it can cover seven buldings in a MetroGrid network for $35,000 in a matter of days while the fiber deployment over the same area will cost $180,000 and take months to install. With a single customer per building and a single T1 replacement at $500 per month, the payback is 10 months, while a more realistic scenario is three customers per building paying $750 per month for a 10Mbps service for an ROI of two months, the company says.

But it will take more than superficially strong comparisons with fiber to reawaken confidence in light-based technologies. FSO is estimated to account for only 1% of the global wireless industry, even though it is part of development programs at a few major operators, such as Qwest and British Telecom. However, the most prominent specialist, Lightpointe, has had its own financial turbulence, and another pioneer, Terabeam, exited the market to focus on the Wi-Fi and WiMAX products of its latest acquisition, Proxim.

China may be a market for FSO, and its leading equipment maker, Huawei, has OEMd Lightpointe’s systems to use in its Optix range, which it sells in Europe, Africa and Asia. Other Chinese companies also have their eyes on this market, and the Tsinghua Tongfang Research Center has developed an FSO last mile solution. Chinese companies predict that the FSO industry will grow to reach $3bn by next year, although high upfront costs – for manufacturers as well as adopters – remain a huge barrier. But the likely demise of Clearmesh suggests that FSO, for all its promise, will remain a niche.