Optical frequency-division multiplexing offers new alternatives to carriers looking to boost bandwidth.
Carriers are finding it increasingly difficult to stay ahead of the demand for higher-rate services. This is especially true with the influx of voice, data, and video driven by Internet traffic in the marketplace. Carriers are constantly looking for solutions that will maximize capacity in their network while meeting such needs as low-cost, flexibility, scalability, and interoperability. These issues have never been so critical as they currently are for the metropolitan marketplace.
In the long-distance market, dense wavelength-division multiplexing (DWDM) has proven the best solution for handling this dynamically increasing traffic. DWDM offers the long-distance carrier such features as capacity expansion, protocol and data-rate independence, and scalability that eliminates the costly need to pull new fiber. Given the success of DWDM for long-haul applications, it's only natural to look at it for the metro market.
In addition to a number of variations of DWDM currently being touted for the metro market, there is a new technology just introduced called optical frequency-division multiplexing (OFDM). Both technologies claim to solve the metro bandwidth problem and both are vying for the attention of the local-exchange carriers.
Carriers in the metro market have a specific set of needs that must be met in order to provide both the quality and the quantity of services demanded by their customers. As a preliminary matter, any bandwidth solution must have a low first cost, which allows carriers to increase the capacity of the network incrementally in response to customer demand. In turn, this enables a short-term return on investment from new revenue streams to pay for the capitalized infrastructure improvements.
Concurrent with the low-first-cost requirement is the need to keep life cycle costs down, which involves more than simply providing reliable equipment. It means ensuring that the solution proposed involves ease of engineering, operations, and training.
In terms of functionality, the bandwidth solution must have a high degree of scalability. It must be able to grow with the network, rather than be hamstrung by design or engineering deficiencies. Given the rate of advance in high technology generally and in telecommunications in particular, carriers are wary of installing equipment that does not have a substantial measure of futureproofing built in.
The metro environment is characterized by rapid and complex traffic moving between closely spaced central offices. Thus flexibility, as in the ability to add and drop individual signals at any central office along the network, is an essential feature of an effective bandwidth solution. As seen in Figure 1, flexible add/drop multiplexing allows new information channels to be remotely added to or dropped from a high-speed signal at each node along the path of an optical signal.
Finally, one of the most critical features sought after in the metro market is interoperability. It makes little sense for a carrier to invest in equipment that performs brilliantly in isolation but does not play well with existing equipment in their network. Carriers are therefore looking for a solution that works with other equipment like Synchronous Optical Network (SONET), legacy equipment, Asynchronous Transfer Mode, and Internet protocol.
Besides equipment interoperability, there must also be a measure of operations interoperability, in which the bandwidth solution fits into the existing operations infrastructure. Thus, managing the ideal broadband network should be the same as managing the existing network.
Given the success that DWDM has enjoyed in long-haul applications, there is a certain degree of appeal when considering it for the metro market. Over the long haul, pulling fiber is not often a realistic solution, and DWDM met this challenge successfully. DWDM technology offers large-scale capacity expansion with the added advantages of protocol and data-rate independence, as well as scalability. As a completely optical system, DWDM has a transparency that enables it to carry just about any signal over light. There are, however, a few areas where DWDM falls short.
High cost: One of the most noticeable aspects of DWDM is its cost. DWDM systems are expensive to purchase, install, and maintain in metropolitan areas. In a long-distance system, several optical amplifiers are often used along the length of the fiber between distant nodes. DWDM increases the number of channels that can share a single optical amplifier and a single optical fiber. This approach substantially drives down the cost of the long-distance system when compared to the alternative of using several fibers, each with its own string of optical amplifiers.
Unfortunately, this cost advantage does not hold for the metro market, since the nodes are much closer together and optical amplification is usually not necessary. Ironically, use of DWDM in a metropolitan setting often requires the addition of optical amplification that otherwise would not be necessary. This amplification is required to overcome insertion losses of the DWDM equipment.
Minimal flexibility: While some metro carriers have selected DWDM, it has only been in limited deployment and typically in very large backbone nodes (multiple OC-48 inputs) in the interoffice environment. Long-haul DWDM systems include expensive optoelectronic conversion on each wavelength at each node, which is cost-prohibitive in this space. Metromarketed DWDM systems, those without such optoelectronic conversion, lack many of the flexibility functions required for metropolitan networks. The filtration of wavelengths in this process produces a loss budget from node to node in which the signal-to-noise ratio (SNR) is increasingly diminished. Such systems lack the flexible add/drop and remote configuration features needed to manage dynamic metropolitan traffic patterns. Networks must be carefully laid out in advance, with planners devoting much of their resources to predicting precisely how metro traffic patterns will behave.
Optoelectronic conversion is also necessary for effective fault isolation and performance monitoring. Metro DWDM systems which use optical filters to pass through wavelengths at a node have no way to provide performance monitoring on those pass-through wavelengths. Thus, if degradation were to occur on any of those pass-through channels in the network, carriers have no way of isolating the fault between offices. Furthermore, the lack of electrical regeneration at each node makes the network harder to engineer, since network planners must be concerned with accumulated jitter and SNR degradation as the number of nodes in the network increases. Thus, most metro DWDM systems are limited in the number of nodes in a ring that they can support as well as the total circumference of the ring.
Competitive local-ex change carriers and other metro carriers are then left with a kind of Faustian deal-to get the bandwidth they need, they have to settle for a slightly lesser degree of functionality. To some, the bandwidth and transparency advantages are worth it.
The architecture of OFDM is based on the principle of technology agnosticism. There is no dogmatic adherence to any one technology or methodology, only a keen awareness of the needs of the market. OFDM was conceived from a broad range of sources, as diverse as aerospace and satellite technologies. It thus emerges as a unique synthesis of three proven technologies-FDM, digital signal processing (DSP), and optical modulation. As shown in Figure 2, it is a multiplexing technique that transports many channels on a single optical carrier, or wavelength.
How it works: With OFDM, multiple channels at different frequencies are combined electrically, and the resulting composite signal is used to modulate the intensity of the light from a single laser transmitter. The optical signal is received at the far end of an optical fiber by a single photodetector that converts the optical signal to an electrical signal. From this point on, the signals are processed electronically. The group of channels is separated into constituent signals by a sequence of frequency conversions and electronic filters. These operations are accomplished with standard components commonly found in consumer electronics, such as digital cable modems and direct-broadcast TV systems.
OFDM offers the ability to transport signals in a protocol- and bit-rate-independent fashion and deliver the bandwidth required by metro markets. Its use of a "single fiber, single wavelength" solution enables multiple signals to be sent down a single fiber-optic cable, expanding metro network capacity, as does DWDM. OFDM, however, adds the dynamic and comprehensive features that networks need, which can be used alone or in combination with existing DWDM systems.
Affordability and reliability: OFDM's combination of proven technologies provides a variety of advantages, and does so at substantial savings. Although the streamlined simplicity of the architecture delivers considerable cost benefits, there is also the advantage of economies of scale. The components have been in use in the consumer electronics industry for years. Therefore, there is no hidden optical R&D cost to sneak into the pricing structure of an OFDM product. Thus, OFDM delivers a low first cost to the carrier.
Life cycle costs are also substantially reduced while efficiency is enhanced. OFDM technology is operated and managed much like SONET systems. There is no need to train technicians to use optical FDM equipment, since the methodology is already second nature to network managers. OFDM also allows network managers to conserve their engineering resources by being able to check the pulse of their network at each node. Since the incoming signal undergoes 3R regeneration at each node, full performance monitoring is conducted on all channels, even for pass-through channels.
By using OFDM, network planners enjoy further design benefits because there is no need for concern regarding accumulated jitter or SNR degradation as a function of the network's size, thus simplifying network engineering and planning. There is no limit to the number of nodes that can be placed on a ring.
True flexibility: Since all channels are regenerated using a single optoelectronic conversion, then pass-through, adding, and dropping can be accomplished on the fly. The node limitations of signal loss and location are no longer a factor. The OFDM waveform undergoes full 3R regeneration at each network node. These factors ensure that the customer can tailor what the network looks like, as opposed to being stuck with the laws of optical physics.
Another benefit to this structure is that it is actually less expensive than alternative technologies. Since there is only one optoelectronic conversion, and there is no need for separate optics, this level of functionality can be maintained very cost-effectively.
Bandwidth efficiency: OFDM also makes conservative use of the limited bandwidth environment. It delivers 10 Gbits/sec in significantly less than 20 GHz of spectrum, maximizing the use of the available spectrum while at the same time delivering greater capacity. The core technology can then scale beyond 10 Gbits/sec either via OFDM or a combination of OFDM and DWDM to gain additional bandwidth cost-effectively.
Full scalability: OFDM returns a certain degree of fair valuation to services by scaling on both the low-speed tributary side and high-speed multiplexing side. This means that increasing capacity from 2.5 Gbits/sec to 10 Gbits/sec is achieved incrementally by adding the necessary capacity on both the low-speed and high-speed sides. Initial costs aside, customers can use the capacity they need as they are obtaining revenue from services, as opposed to paying for the full high-speed multiplexing before they actually need it.
Dynamic interoperability: OFDM shows the hallmarks of a truly dynamic technology. It maintains all the transparency advantages of optical modulation, such as carrying anything from circuit-based traffic to high-capacity data to signals from legacy systems, but also produces a signal that is extremely robust against dispersion and fiber nonlinearities. OFDM demonstrates a healthy degree of fiber-quality insensitivity, maintaining signal quality regardless of the vintage of the fiber. Furthermore, unlike DWDM, OFDM excels on dispersion-shifted fiber. While this last feature may be unimportant in the United States, carriers in countries like Japan will find this aspect valuable in solving their bandwidth issues.
One of the critical features of the ideal broadband solution is that it be fully interoperable with the existing range of equipment and protocols. OFDM is complementary to both DWDM and SONET, and enhances both, resulting in a network that provides the advantages of each technology.
For example, for those carriers that have already installed a DWDM system, adding an OFDM component would confer all of the enhanced functionality and interoperability mentioned above into the metro DWDM architecture. Such a combined system could enable the transmission of a bandwidth-efficient composite OFDM waveform in each of the wavelengths of a DWDM system. The bandwidth potential of such a combination system would bring together the most effective and powerful aspects of the state of the art in multiplexing technology.
Carriers are keeping an eye on the future. As technology continues to evolve to keep up with the massive growth of Internet traffic, they are seeing possibilities with OFDM as both an alternative and complement to existing equipment.
Multiplexing research has been fairly optics-oriented, which is understandable in light of the phenomenal benefits in bandwidth and transparency. There is also a substantial benefit from taking an "interdisciplinary" approach to technology, such as folding DSP and FDM into optical modulation. The result of such an approach is that OFDM stands ready to deliver, either by itself or in conjunction with existing legacy systems, the complete range of features demanded in the metro market.
Michael Rowan is CEO and president of Kestrel Solutions (Mountain View, CA).