Advancing the Optical Component
telecoms-mag.com
There is no doubt that the widely used and well-defined synchronous optical network/synchronous digital hierarchy (Sonet/SDH) technologies and their applications will continue to evolve. Key advances are now being made in the optical component arena to make the photonic vision a reality.
Andrew Hunwicks
--------------------------------------------------------------------------------
In the first few years of this millennium, there will be an explosive growth in the amount of information being transmitted by digital services. Electronic commerce, software distribution and digital video/music distribution services are just a few of the applications in the emerging data-centric communications environment, involving multimedia networks based on Internet protocol (IP) and to a lesser extent, asynchronous transfer mode (ATM). These trends focus attention on the need for more capacity, spurring on the advent of new, more efficient next generation communication technologies.
In this competitive market, photonic networks, based on wavelength division multiplexing (WDM), offer a dynamic combination: the ability to increase the communication capacity while at the same time drastically decreasing the transmission cost per channel. The technology can radically improve node throughput, offering transparency and flexibility while accommodating different signal formats and protocols. Expandability is inherent to the photonic concept, allowing for a cost effective gradual deployment in line with network demands. These features are combined with high reliability and ease of management to make up a compelling business and technological argument for photonic network implementation. The telecoms industry is moving to a world of petabit, terabit and gigabit ranges, as opposed to the current provision of gigabit capacity in the core backbone network with megabit access for enterprise and dedicated networks, and mere kilobit capacities for residential usage.
Behind the Photonic Scene
Essentially, a photonic network allows many high bit rate channels to be multiplexed over a single fibre strand -- namely, dense wavelength division multiplexing (DWDM). While not in itself a new technique (early trials of multi-channel WDM were performed as long ago as 1988) the practical realisation has only recently become a reality. The reasons for this long gestation period are linked with the availability of reliable and robust optical components. So why are network operators looking to deploy photonic networks? The answer lies in two key areas -- firstly, the growth in services based upon IP, which in order to offer high quality of service, requires very high bandwidth connections between nodes. Secondly, there is the requirement for flexibility within the network architecture, which allows operators to reconfigure and reallocate bandwidth as market demand changes. Combine these with the ability to reduce operational and maintenance costs, as well as the costs associated with network build, and you have a compelling business argument.
Photonic Network Components
The development of photonic networks has already moved through the initial phase of simple point-to-point WDM links to the second phase in which optical add/drop multiplexers (OADM) are now commercially available. This key network element is designed to make efficient use of network capacity, network protection, wavelength routing and to conduct the majority of the optical network?s functions. In specific terms, the OADM site is where the DWDM optical channels are generated and amplified (Figure 1).
The high functionality of the OADM node in turn requires a number of optical subsystems, many of which have only recently become available. At present, there is a range of OADM technologies on offer, encompassing totally passive configurations to the latest dynamic solutions. The traditional passive configuration utilises thin films interference filters, fibre gratings and planar waveguides. Optical characteristics, such as insertion loss, plus interband and intraband crosstalks, are well defined for each of these technologies when used in a passive OADM application. However, dynamic OADM nodes provide other more flexible possibilities. Dynamic OADMs have the major advantage of being able to deliver improved cost effectiveness and flexibility compared to passive OADMs. This is achieved by the ability to select and manipulate any wavelength channel, thereby provisioning circuits on demand without changing the physical configuration. So, there is a strong argument to ensure that a smooth migration from passive to totally reconfigurable and dynamic OADM is achieved. In order to realise this goal, new components such as acousto optical tunable filters (AOTF), tuneable lasers and wide spectrum optical amplifiers will all be required. With the requisite micro-electronics skills, these remarkable technologies can now be transferred from the development laboratory into today?s state-of-the-art network elements.
Towards an Integrated AOTF
As an example of the type of component used in today?s dynamic OADMs, we should consider the AOTF in more detail. In this key component, the incident light is propagated over the optical waveguide and divided into perpendicular components Transverse Electric/Transverse Magnetic (TE/TM) by a perpendicular beam splitter (PBS). An acoustic wave is generated by applying a RF signal to the inter digital transducers (IDT). This acoustic wave travels through the SAW guide and causes a periodic modulation of the refractive index of the optical waveguide. This change of refractive index induces TE-TM or TM-TE conversion for only the drop wavelength. The drop wavelength corresponds to the applied RF frequency and becomes perpendicular to the incident light. The second PBS is then used to separate the drop wavelength from the incident light. An AOTF can be used to drop a single wavelength or multiple wavelengths in a simultaneous manner. By changing the number of RF signals and their respective frequencies, it is possible to control the number and frequencies of the drop wavelengths. There are no moving parts in the AOTF, making it extremely reliable, and it offers high-speed wavelength tuning that can be performed sequentially or randomly on the applied RF frequency. Whilst AOTF devices offer a number of significant advantages, there were several issues associated with previous generation AOTF devices, such as high insertion loss and poor sidelobe suppression. However, recent technical advances have significantly improved both of these characteristics. In Fujitsu?s laboratories, for example, this has involved the development of the intersecting-waveguide polarisation beam splitter (PBS), a new film-loaded SAW guide (SAWG) and a 0-gap directional coupler type reflector, which has contributed to improved performance for the AOTF and enabled the integration of an AOTF into a single chip.
Tunable Laser Developments
A second key component for fully flexible OADMs is the tuneable laser. Currently, narrow line width lasers can be created using a variety of techniques such as the following: distributed brag reflector (DBR) lasers; external cavity diode lasers (ECDL); vertical cavity surface emitting lasers (VCSEL); or distributed feedback (DFB) lasers. However, when used in DWDM applications these discrete components are not ideal because separate devices are required for each WDM wavelength. The advent of tuneable lasers avoids this limitation and now enables a fully dynamic add/drop optical architecture to be constructed. This network architecture, using combinations of tuneable laser and AOTF, can realise wavelength provisioning services, broadcast network topologies, wavelength converters, optical cross connects and wavelength routers. It also represents a means of substantially reducing the costs associated with AODM elements at the internal testing stage -- a major consideration given that AODMs consist of costly subsystems and complex manufacturing processes. Once the element is in a live network environment, the tunable device can be used for the following: spares in the cold stand by mode; restoration in hot stand by; rerouting wavelengths; fast packet/OXC/all-optical network; or wavelength conversion.
In order to see a wider commercial deployment of tunable lasers, there are some other challenges that have to be addressed. Tunable lasers need to demonstrate the following characteristics:
A wide tunability range, preferably for full C bands;
the optical output power of the device is required to be 10dBm or more;
the cost structure is required to be close to DFB lasers;
the device performance requires low laser chirp, high modulation speed, small size and very high reliability;
the device needs to conform to international standards.
A wafer scale approach, offering a low-cost, high performance solution, is definitely one avenue that could address these issues. One integrated technique is the tuneable VCSEL configuration, which offers the advantage of allowing for pre- and post-fabrication testing, as well as wavelength tuning. This laser subsystem is a simple voltage-controlled tuning device, with a wide tuning range. Such devices allow for direct modulation between the 25 Gbps and 10 Gbps range, and additionally, they can be easily adopted within standard telecom packages.
Optical Amplifiers
Moving onto the question of optical amplifiers -- in order to provide the necessary broad bandwidth, low noise and high power for a wider input power range, these devices must provide gain slope control and output level control. For long distance applications, there must also be a compensation mechanism for dispersion. In addition, devices deployed in the field are required to handle channel upgrades, band upgrades and be directionally flexible. The majority of today?s installed base of optical fibre consists of either non-dispersion shifted fibre to specification ITU G.652; dispersion shifted fibre to ITU G.653; or non-zero dispersion shifted fibre to ITU G.655. Most transmission systems have evolved from 1310 nm single channel systems to multi-channel 1550 nm DWDM systems in the C band. The C band is useful as the signals can be amplified by erbium-doped optical fibre amplifiers (EDFA), which are commercially available. The EDFA consists of three main components: a semiconductor pump laser; a length of erbium-doped fibre; and a degree of doping concentration. By tailoring these components within the amplifier, it is possible to extend the amplification window into the L Band. This means that by changing the type of laser pumps and threshold operating power; or by changing the length of erbium fibre; or doping concentration, additional amplification range is achieved.
L band amplifier modules are deployed as a separate unit. The L band wavelengths are separated from C band wavelengths and fed into a separate amplifier, which can be either single stage or dual stage.
L Band amplifiers increase the bandwidth available on a single fibre and consequently offer flexibility and a significant addition to optical network capacity. Despite the costs associated with specialised erbium fibre, filters, couplers and dispersion compensators, the L band amplifier is now cost effective for many applications. While EDFA provides an efficient amplification of the C and L band optical windows, the S band is still largely unexplored. However, many research laboratories, Fujitsu included, are investigating potential amplification methods and recent photonic developments point to the potential advent of S band amplification based upon thallium-based amplifiers.
Looking to the Future
There is no doubt that optical technologies are transforming the way in which future network development is approached. The initial deployments of WDM line systems have successfully demonstrated the commercial advantages to be gained from channelling more traffic through fibre. The challenge now is to demonstrate how this approach can be maximised to provide a revolutionary new approach via all-optical networking. There has been a steady series of optical component breakthroughs en route to the current situation, in which fully flexible OADMs are now available, which allow operators to access optical channels economically at intermediate nodes. This has avoided the expense of having to install back-to-back WDM line systems, and opened the door on a new tranche of photonic potential.
The technology is clearly advancing rapidly and once there is a clear match between what is economically viable and technologically feasible, complete photonic networks will be genuinely viable. En route to this point, metropolitan applications are already demonstrating the relevance of this approach. While the technology challenge has been clearly defined, the market situation is less clear-cut. Operators are under increasing pressure to focus on building out their infrastructure with lower cost, higher-functionality products, ensuring a smooth transition between technologies and total equipment interoperability throughout a hybrid network. The bravest will look at the potential of an all-optical network environment and plan for it now.
Andrew Hunwicks is product management director at Fujitsu Telecommunications Europe Ltd. He can be contacted by email: A.Hunwicks@ftel.co.uk, website: www.ftel.co.uk |
