17 Jun 2002
Despite the declining economy, players in the tunable-laser market are making bold statements about growth. Rebecca Pool talks to industry experts to find out why.
From Opto & Laser Europe December 2001
The telecoms industry may be floundering in the current economic downturn, but the market for wavelength-adjustable transmitters (or tunable lasers) remains intact. In fact, the latest reports from market analysts predict that it will see staggering growth over the next five years. One such analyst, the Yankee Group, US, predicts that today's revenues of USD 50 million will grow to USD 2 billion by 2006.
Jay Patel, senior analyst at the Yankee Group, says that although the current economic climate has had some effect on the tunable-laser market, growth will still occur. "The economic downturn means that we are starting from a lower baseline," he said. "But we are still seeing growth and we are just as optimistic as before."
According to Patel, tunable-laser-market growth is likely to track the ever-increasing demand for greater fibre capacity in fibre-optic networks, whether for long-haul or metropolitan markets. Dense wavelength-division multiplexing (DWDM) technology has increased the number of commercially available transmission channels in fibre-optic networks to 40 wavelengths per fibre, and the technology already exists to raise the number of channels to as many as 172 wavelengths per fibre.
But while this may sound like the ideal solution to data-hungry consumers, the reality is a
logistical nightmare for network operators. Existing DWDM networks typically use a unique laser for each
wavelength in the transmission fibre, and as the number of channels in DWDM networks increases, more
fixed-wavelength lasers are needed. From the service provider's point of view, stocking spare capacity is
complex and costly, unless it uses tunable lasers. It is not
just the big players who are making waves in this relatively new market. Start-ups that have already made
a name for themselves include US-based firms Agility, Bandwidth9, BlueSky Research, Iolon and
Sparkolor. Patel says that Agility, which announced its tunable-laser product in November last year, plans
to increase its laser-manufacturing capacity from this year's 50 000 to 1 million by 2003. While
inventory management and sparing is the main tunable-laser application today, other applications will
follow. As old transmission links are replaced and new ones are installed, it will be possible to use tunable
lasers instead of fixed-wavelength sources in DWDM systems. Jens Buus, technical director of
UK-based consultancy Gayton Photonics, believes that this is an essential step towards ensuring flexibility
in DWDM networks. He also believes, however, that the cost of tunable lasers must first be reduced.
"Sparing is less sensitive to cost issues, but for tunable lasers to be used as replacements, prices need to be
lower," he said. Looking even further
into the future, tunable lasers have the potential to bring new possibilities to network design. The routing of
optical signals to different locations within a network architecture is determined by the signals'
wavelengths. If a tunable laser is employed in the transmitter, it is possible to re-route the wavelength to
any network node. One firm that has incorporated tunable lasers into its WDM platform for this
reason is US-based Atoga Systems. Currently trialling Agility's 40-channel tunable laser in its systems,
Atoga's vice-president P G Menon believes that these lasers are essential to bringing "intelligence" to a
network. "If, for example, the traffic demand in a network changes, tunable lasers can alter their
wavelength to match the demand," he said. "This is a real benefit to service providers, because it means
that they don't have to physically do it themselves." Menon also says that following laboratory and
field trials of his firm's DWDM platform, vendors that were sceptical about tunable lasers 18 months ago
are now happy with their results. Unfazed by telecoms woes, Menon is confident that Atoga's platform
will be ready for commercial use by the second half of next year. "[The downturn] has only delayed
the installation of our systems until the next quarter," he said. "We have spent months working alongside
vendors to test tunable-laser characteristics and our results have shown them to be robust." Tunable
lasers are also likely to be used in packet switching and they could lead the way to the development of
all-optical networks. Microelectromechanical systems (MEMS), which switch wavelengths from one fibre
to the next, are already emerging. However, if widely tunable lasers are added to MEMS, switching will be
feasible from any wavelength to any other. Service providers are also concerned about
power issues surrounding tunable lasers. "Current power levels are fine for metropolitan communications,
but not for long haul," said Patel. "We could have tunable lasers in long-haul markets by next year, but that
is optimistic." Both Buus and Patel agree that standardization between different laser technologies is
not a problem, although Buus emphasizes that control and reliability need some work. However, he added:
"These issues are resolvable - it is just a matter of time. There is nothing that is intrinsically
impossible." US firm ADC, which took over Swedish company Altitun in May 2000, believes that its
grating-coupled sampled-reflector (GCSR) lasers are the answer. These lasers were first
commercialized by Altitun, which combined a widely tunable co-directional coupler with a sampled grating
to give a laser with wide tuning and a good spectral selectivity. When ADC bought Altitun to form ADC
Photonics, ADC's senior vice-president cited the GCSR laser as a pioneering technology with the widest
tuning range commercially available. Björn Broberg, director of engineering at ADC Photonics
agrees. "Historically we have tried [two other] tunable lasing technologies," he said. "We found that the
GCSR design had a power advantage, and that the power was uniform over the tuning range." In addition,
Broberg says that the technology is robust and easy to control, with a comparable processing and chip cost
to other tunable sources. ADC believes that its tunable-laser product has been able to achieve nanosecond
switching speeds that are fast enough for packet switching. Another tunable-laser technology that
looks set to take a key role in the market is based on vertical-cavity surface-emitting lasers (VCSELs).
VCSELs can be converted into tunable lasers by suspending a top reflector over the active region that
electrostatically deflects to form a small cavity. Tuning ranges are typically between 30 and 50 nm, and the
power output is limited. Both Bandwidth9 and Nortel Networks are developing VCSEL-based
tunable lasers. Having bought CoreTek in the US, Nortel expects to launch its VCSEL technology by the
end of this year. "For long-haul networks, external-cavity lasers will lead because of their
high power, whereas distributed Bragg reflector lasers and VCSELs have low power, which is more
suitable for the metro market," he predicted. However, he adds: "Cost will be an issue, so in the long term,
VCSELs could lead."
Distributed-feedback (DFB) lasers are based on thermal tuning. As well as having narrow tuning ranges and slow switching speeds, these lasers also have a low output power. Distributed Bragg reflector (DBR) lasers, based on simple DFBs, can be either thermally or electronically tuned and, unlike their predecessors, have the potential to reach nanosecond tuning speeds. US-based Agility has taken DBR technology one step further to produce thermally tuned sampled-grating DBR lasers. Meanwhile, Marconi Optical Components of the UK is also working on advanced DBRs and has just announced its digital supermode indium phosphide DBR laser. The laser has a minimum tuning range of 35 nm and Marconi hopes to make volume shipments by mid-2002. Mechanically tuned lasers include external-cavity lasers (ECLs) and vertical-cavity surface-emitting lasers (VCSELs), which typically use MEMS to switch wavelengths. ECLs contain an external grating that acts as a reflector. By rotating the grating, continuous wide tuning up to theoretical values of 240 nm is possible. Key ECL developers include US-based New Focus, Blue Sky Research and Iolon, which has just developed an ECL that is tuned with silicon MEMS. Researchers at Iolon say that by using MEMS they can reduce the device size to a small robust package suitable for use in optical networks. Their ECL offers wide 38 nm tuning ranges and high power outputs suitable for telecoms.
© 2024 SPIE Europe |
|