Project Oxygen poised for construction under new strategy
By ROBERT PEASE
Fiber-optic technology is advancing so rapidly that, in less than a year`s time, one proposed global network has seen its cable requirements cut in half while its total capacity has increased.
In late 1997, ctr Group Ltd. (Woodcliff, NJ) proposed Project Oxygen, the most ambitious global fiber-optic network to date (see Lightwave, October 1997, page 1). Originally, the plan called for linking 262 landing points in 175 countries to a network that would deploy eight wavelengths to achieve a 10-Gbit/sec pipe over two fiber pairs for a potential long-haul capacity of 160 Gbits/sec.
Just months later, ctr Group brought the planned network to the table at a meeting with more than 300 telecommunications carriers from 150 countries. Already the network had undergone subtle, but significant, changes (see Lightwave, February 1998, page 15). The updated plan proposed 265 landing points in 171 countries and the deployment of eight wavelengths over four fiber pairs for a total capacity of 320 Gbits/sec, doubling the original long-haul capacity. Thirty carriers signed memoranda of understanding at the meeting, indicating they were ready to proceed and willing to financially commit to the network.
Technology continued to get the upper hand, however, and the latest changes to Project Oxygen reflect its impact, according to Neil Tagare, chairman and chief executive at ctr. He says newer technology is enabling Oxygen to move quickly to the construction phase and to do more with less.
"Today, we can generate much more capacity on less cable," says Tagare. "Because of the added capacity, we have reduced the core of the network to 158,000 km for the first phase. That`s about one-half the size of the original planned network. We want to begin building the cable as soon as possible, with construction beginning in the first quarter of 1999."
Dense wavelength-division multiplexing (dwdm) technology will allow the network to again double the capacity for long-haul submarine routes. Each cable will contain four fiber pairs on long-haul (at least 350-km) undersea segments, which require optical amplifiers, and 12 fiber pairs on terrestrial and shorter submarine segments. Using dwdm will allow each fiber pair to carry 16 wavelengths, each transporting 10 Gbits/sec for a total of 160 Gbits/sec per fiber pair. Total throughput will be 640 Gbits/sec for long-haul submarine routes and 1.92 Tbits/sec for terrestrial and shorter undersea segments.
Phased construction
Another major change to Oxygen involves a phased construction schedule. Phase One is planned for completion in early 2002 and will be constructed in four parts (see figure on page 1). Major transatlantic and transpacific routes, as well as a terrestrial link across North America, will be completed in late 2000. Northern Europe will be connected in early 2001. In late 2001, the Mediterranean, the Middle East, India, and both coasts of South America will be connected. Finally, Australia and Central America will be linked in early 2002, completing the initial phase. A subsequent second phase will increase regional connectivity to the core network. The network will be built in a loop configuration rather than point-to-point. Originally planned for 38 loops, it will now require just 17 to provide comparable capacity.
Startup funding for the project is being provided by Tyco Submarine Systems Ltd. and Corning Inc. of the United States; Alcatel Submarine Networks of France; and nec Corp., ntt International Corp., Sumitomo Corp., and Mitsui & Co. Ltd. of Japan. These companies are not equity owners of either ctr Group or Project Oxygen Ltd. Their funding provides them with the first opportunity to supply the equipment that will make up the Oxygen network.
Project Oxygen was originally planned as a consortium cable, with telecommunications carriers financing the entire project and becoming joint owners with the right to use and sell excess capacity on the system. However, it is now being structured as a private system, says Tagare, because of the substantial interest in the project from private companies wishing to invest. As a private system, Oxygen`s financing will come from both carriers purchasing excess capacity and private lenders and investors providing debt and equity. The unsubscribed capacity will be sold to service the debt and provide returns to the equity investors. The financial closing date is scheduled for December 1, 1998.
Although the agreements for most equity partners, vendors, suppliers, and contractors are still in negotiations and may be several months away, at press time ctr had announced two major equipment contributors to the project. Corning Inc. (Corning, NY) will supply optical fiber for the cable and Lucent Technologies (Murray Hill, NJ) will provide land-based optical networking equipment, network management, power systems, and interconnect equipment.
Making bandwidth less expensive for carriers is a critical element of the project, says John Beagley, vice president of Lucent`s Optical Networking Group. One contribution Lucent will make toward that goal is the use of its WaveStar Optical Line System 400G, a scalable dwdm product that can increase capacity on a single fiber strand up to 80 wavelengths. But both Tagare and Beagley agree the most significant contribution Lucent will make to Project Oxygen is in network bandwidth management.
"We`ll deploy the WaveStar BandWidth Manager on this network," says Beagley. "Basically, this system enables management of various types of fiber connections coming into the system, routing voice, Asynchronous Transfer Mode, Internet protocol, and voice traffic over the same network to all the global locations."
Beagley says the Oxygen network is a "great fit" for the products because they allow management of massive amounts of bandwidth and are fully scalable for future technologies.
"Project Oxygen is going to be selling the reliability aspect of this network," says Beagley, "because so much bandwidth will be traveling across it. We`ll provide a ring protection feature that we`re currently patenting, which was developed by Bell Labs."
The network will be built and serviced under a supply contract between Project Oxygen Ltd. and its selected suppliers. Maintenance and repair for the system will be provided by a ctr-affiliated company, ctr Maintenance Service Co. The three network management centers will be located in the United States, Spain, and Singapore. Total construction cost for Phase One is $8 billion. q