Coincident with rate and signal density increases of the past few years, a shift has occurred in the nature of traffic transported on backbone networks. Today's traffic is data-centric which, when coupled with technology advances, divide the long-haul space into long-haul and ultra-long-haul networks.
Ultra-long-haul networks provide express pipes between major nodes on the network, extending transmission reaches beyond 1500km and reducing costly signal regeneration. A variety of techniques, such as optimised fibre profiles, Raman amplification, dispersion and slope compensation, forward-error-correction and advanced signal modulation, are used to achieve these distances and data rates. Traditional long-haul networks focus on traffic aggregation and regional transport, where the distance between major nodes is less critical.
For long-haul networks, advanced technologies can be applied to increase overall system margin. As the percentage of data traffic in a network continues to increase, the emphasis on ultra-long-haul transport will also increase.
Network providers are primarily focused on two things: lowering costs and increasing revenue. A benefit of optimised fibre design is the ability to reduce network costs, both capex and opex.
Optimised fibre designs provide network designers greater system margin. This margin can be utilised either to increase system reach or relax component specifications for equivalent reaches. Increased system reach reduces the amount of regeneration equipment in the network, while relaxed component specifications lower costs. Also, optimised fibres allow network engineers to increase the system's spectral efficiency, which in turn maximises the available information-carrying capacity for a given wavelength range. And lower costs are achieved by eliminating or reducing the number of components.
To increase reach while maintaining total bandwidth requires a new fibre solution both to increase performance and provide more uniform performance for all wavelengths. The former can be achieved by decreasing both the amplifier noise figures and total span loss, while the latter is achieved by more accurate broadband dispersion control.
There are two ways to increase spectral efficiency: tighter channel spacing, such as moving from 100GHz to 50GHz or even 25GHz; and raising per-channel bit rates (say, from 10Gbit/s to 40Gbit/s). Increasing data rates requires precise dispersion control from the fibre. The primary foci for component reduction or elimination are expensive per-channel devices, such as dynamic dispersion compensating modules and polarisation mode dispersion (PMD) compensating modules. The fibre requirements to achieve these are precise dispersion control, dispersion stability and lower PMD.
When designing an optical fibre, its optical profile significantly impacts three major attributes: chromatic dispersion, dispersion slope and effective area (Fig.1). The profile, in theory, can be manipulated to optimise independently any one of these attributes.
However, these properties are inter-dependent, making simultaneous achievement of all of them virtually impossible. For instance, lowering the dispersion slope also lowers the effective area. Historically, fibre design sought to balance these parameters to meet commercially available system requirements. In order to meet the demands of today's and tomorrow's challenging market conditions, networks will need to progress beyond current performance and cost capabilities.
Dispersion-managed fibre (DMF) is a relatively new concept in which two or more fibres are combined within a transmission span — thus achieving optimisation of dispersion, slope and effective area simultaneously. Early versions of dispersion management implemented in submarine systems consisted of combining positive-dispersion, positive-slope fibres with negative-dispersion, positive-slope fibres to achieve a flatter dispersion profile. But using two positive-slope fibres still leaves an overall net positive slope on the transmission link.
It wasn't until fibre designers developed an acceptable negative-dispersion, negative-slope fibre that the value of dispersion management was realised. This technology may now be applied in terrestrial ultra-long-haul networks with advantages.
For a terrestrial DMF solution, imagine a span of fibre between amplifier sites, typically 80–100km apart in a terrestrial network. The first segment of the span is a large-effective-area, positive-dispersion, and positive-dispersion-slope fibre. The middle segment transitions to a smaller-effective-area, negative-dispersion, and negative-dispersion-slope fibre. The final segment mirrors the first.
A pulse travels along the span and broadens very rapidly in the first section, due to the positive dispersion. As the pulse travels within the negative-dispersion segment, it compresses — reversing the broadening effect. The length of this segment is such that the pulse actually compresses beyond its original state. In the final segment of positive dispersion, the ultra-compressed pulse again broadens, finishing the span in its original state (Fig. 2). With excellent matching of both dispersion and dispersion slope across the C and L bands, the end result is not only a single non-dispersed signal pulse, but flat dispersion across the entire operating window. Deploying a DMF solution can increase overall system because of many factors (see Table).
Reduction of network components is a result of two primary factors. First, by combining the proper ratio and lengths of positive-dispersion and negative-dispersion fibres, it is possible to achieve broadband dispersion compensation in the outside plant, rather than with additional dispersion compensation modules. Secondly, the broadband dispersion control reduces the dynamic range or number of tunable compensation devices required for high-bit-rate systems.
Implementing dispersion-managed fibre solutions in terrestrial applications requires more than just a superior optical fibre — it will also require concurrent advances in technology — along the entire optical chain from photonics to outside plant practices.
Understanding the requirements placed on the outside plant (OSP) will be an integral part of deploying next-generation, high-data-rate, long-haul and ultra-long-haul networks (Fig. 3).
Fibre manufacturers such as Corning will continue to research new technologies that deliver value to the telecommunications industry. Whatever the market conditions, constant trends are bandwidth and customer growth. The global economy demands more bandwidth-intensive applications to keep up with the technology and competition. To sustain this growth, while remaining competitive, carriers will need to combine advanced transport solutions with optimised fibre designs.
Joel A Orban
Manager, New Products
Corning Optical Fibre
Joel Orban is responsible for identifying and commer-cialising new optical fibre products to address high-rate terrestrial and metropolitan network needs.
- The large effective area of the positive-dispersion fibre allows higher launch powers while suppressing non-linear penalties.
- The high local dispersion of the positive-dispersion fibre suppresses non-linear penalties, while the overall near-zero.residual dispersion of a span optimises system performance.
- Eliminating the need for dispersion compensation modules reduces the noise figure for erbium-doped fibre amplifiers (EDFAs) by decreasing the amount of loss for the EDFA to compensate.
- Reducing or eliminating dispersion compensation modules reduces their contribution to non-linear penalties.
- The small effective area of the negative-dispersion fibre is an efficient Raman amplification gain medium. Including it in the actual transmission line allows further distribution of Raman gain as well as lower overall noise.
- The complementary design of the positive- and negative-dispersion fibres provides less variation in dispersion with temperature fluctuations.