Recent advances in optical technology are rapidly making possible something that just a couple of years ago was considered unthinkable: The deployment of fiber to nodes so small that the network has become almost entirely passive, and fiber is pushed so close to the home that consumers can nearly reach out and touch it.
As cable operators prepare their networks to carry not only interactive video, but telephony and data as well, they're bulking up their networks with ever-greater amounts of bandwidth to handle all that traffic. What they've discovered lately is that technical innovation and precipitous price drops in the optical world are combining to make it possible to drive fiber into nodes of 150 homes or less.
Leading the charge are the country's largest MSOs, who are focused on laying the groundwork for new, interactive services by upgrading, in particular, the return portion of their networks. "Everyone is definitely focused on bringing fiber closer to the home," says Emmanuel Vella, vice president of product management for Antec's Active Electronics Group. "Several operators and overbuilders are very aggressive and are being driven by the proposed addition of two-way services, including telephony, as well as the enhanced reliability that fiber brings."
"Fiber to the 100- to 200-home (node) level is a given," agrees John Trail, manager of transmitter systems at Harmonic. "The amount of bandwidth required is rising and fiber deeper gets you more capability and capacity." Trail notes that the previous benchmark of 600-home nodes simply doesn't offer enough capacity for an operator to offer high-speed data and telephony.Real-world application
AT&T Cable Services is currently testing a concept it calls the "LightWire neighborhood broadband system" in about 700 miles of plant in Salt Lake City, according to Tony Werner, AT&T Cable Service's executive vice president and chief technology officer. The system was jointly developed by Werner's group and engineers at AT&T Labs. When AT&T purchased Tele-Communications Inc., the concept of the mini fiber node, which had been developed by AT&T Labs, was presented to Werner. "We continued to find fault with the architecture," says Werner, smiling. "Certain aspects didn't make sense, but certain things made a lot of sense. Over time, it became a pretty attractive architecture."
By that, Werner means that it offers the bandwidth the company needs, yet appears to offer tremendous operational cost savings, including powering and maintenance costs.
The LightWire architecture essentially replaces traditional coaxial amplifiers with "mini fiber nodes" that were designed by AT&T Labs researchers. These mini nodes perform the same function that traditional nodes do in that they convert signals from the optical domain to electrical (and vice versa in the upstream direction). These nodes, which are housed in bridger-sized packages, can often replace as many as two or three conventional RF amplifiers. The network from the mini nodes to the customer premise becomes entirely passive.
A new device, dubbed the MuxNode, terminates the fibers that connect multiple mini fiber nodes. It multiplexes upstream signals and sends them back to the primary hub locations via the secondary hubs. It also demultiplexes downstream signals that come from the primary and secondary hubs and distributes them to the mini nodes. (For complete technical details of the AT&T LightWire architecture, see "HFC architecture in the making," July 1999 CED, p. 58.)
With the architecture, AT&T engineers hope to increase the upstream capacity of the network by a factor of 10 or more, while reducing the number of active components in the network by as much as 60 to 80 percent. This reduces overall cost while dramatically increasing reliability levels, according to Werner. An additional benefit is an anticipated 50 percent drop in powering costs as well.
The Salt Lake City test of the system is presently being planned and should be operational before the end of 1999, Werner says. Shortly after that, AT&T should have the economic and performance data it needs to determine if the architecture can be implemented elsewhere. Werner believes that if the test goes well (which he anticipates), the system will be deployed nearly everywhere going forward. This "passive" approach of delivering fiber deeper into the network is also being touted by a Davis, Calif.-based company called Alloptic Fiberoptic Access Networks. Alloptic executives say its fault-tolerant, dual-ring passive optical network (PON) can be dropped in between the headend and the fiber nodes for roughly one-fifth the cost of a Sonet-based network.
(Details on PON architectures and how they can be applied to HFC networks will be provided in an upcoming issue of CED.)Telcos in the act, too
And the cable industry isn't alone in driving fiber ever closer to the home. BellSouth has marshaled partners both domestically and abroad to advance fiber technologies. The Atlanta-based telco currently has three separate fiber initiatives underway, all designed to maximize fiber use cost-effectively. According to John Goldman, communications manager for BellSouth, the company began deploying fiber-to-the-curb (FTTC) about three years ago in newbuild and new development areas "all over our service area." In fact, Goldman estimates that more than 200,000 homes are now served by FTTC, and by the end of the year, that number will jump to 300,000.
Beyond that, at the beginning of this year, BellSouth began an "integrated fiber in the loop" (IFITL) installation program. IFITL is a network rebuild initiative that integrates copper and fiber (and coax in areas where the company has cable franchises). The architecture is designed to provide easy migration to fiber-to-the-home. This year, fiber will be installed to within 500 feet of 200,000 homes in the greater Atlanta and Miami areas, replacing existing copper twisted pair.
Customers served by IFITL deployments will enjoy the same services they've always received, including voice, video and data (where available). Asymmetric digital subscriber line (ADSL) service will continue to be offered to IFITL areas, yet the pesky distance limitations of ADSL-over-copper will be eliminated with the addition of fiber, says Goldman. However, BellSouth hopes to deliver ADSL to areas via IFITL which were previously too expensive to serve. "IFITL helps to round out our coverage area," says Goldman.
Marconi Communications Inc., according to Joe Ross, general sales manager, is providing BellSouth with a supervisory software system at the central office as well as host digital terminals which work off the central office terminal. An optical network unit is placed within 500 feet of the home and includes a video coax drop and voice/data twisted pair interface.Fiber-to-the-home
Driving fiber even deeper, BellSouth has been experimenting since June with fiber-to-the-home in 400 homes in Atlanta. The project was announced in June and eventually will deliver voice, video and data services on fiber terminated at the customer premises. The company is overlaying existing copper with fiber in a select Atlanta residential neighborhood to residents who have a high probability of buying data and/or video services. The trial will utilize asynchronous transfer mode passive optical networking (ATM PON), which is believed to be the first use of the technology in North America. BellSouth has a joint research agreement with Japan's Nippon Telegraph and Telephone Corp. to develop FTTH technologies.
The idea behind ATM PON is relatively straightforward-elimination of active electronics requiring AC power that are typically located in pedestals and outside cabinets between BellSouth's central offices and its customers. Only optical fiber and passive optical splitters are used in the outside plant. In BellSouth's configuration, there is a four-way splitter at the curb and an eight-way splitter further up the network, allowing each optical line terminal in the central office to be shared by up to 32 customers, according to Dan Spears, research director in BellSouth's science and technology group.
However, such a scheme depends on local powering at the customer's home to run the optical transmitter/receiver. The addition of ATM to the network provides a way to deliver voice, video and data from a number of sources, multiplexed on a common network and switched to intended destinations.
Lucent Technologies and Oki Electric Industry Co. developed the optical fiber access system used for FTTH deployment, consisting of Lucent's optical network access equipment, Oki optical line terminals, a network operating system and analog optical network terminals. Lucent's optical network termination unit will be placed within a customer's home to convert optical signals into high-speed Ethernet data for the PC. The Oki analog optical networking unit will provide video signals for a customer's TV.Lots of choice
BellSouth hopes to deliver 120 channels of digital video, 70 channels of analog video, 31 channels of CD-quality audio, and high-speed ADSL data service. Lucent is providing the optical access termination unit, which uses high-speed Ethernet and is placed inside the home, while Oki Electric is providing optical line terminals to convert video signals. While ATM PON has the potential to deliver data service at a sizzling 100 Mbps, Goldman says BellSouth will only offer its existing ADSL service (1.5 Mbps upstream; 256 kbps downstream) in its FTTH areas, although that may change in the future.
Goldman points out that FTTH technologies are being driven by the Full Services Access Network (FSAN) industry consortium, a group of 20 telecommunications companies and vendors working to develop specifications for FTTH interfaces and speed FTTH technologies to market. In June, four FSAN telcos-NTT, BellSouth, France Telecom and British Telecommunications plc-issued a common technical specification for ATM subscriber systems. More information about the spec and FSAN may be found at http://www.labs.bt.com/profsoc/access/. Spears expects to see equipment based on the spec in the next few months, and in service by the end of the year. The FSAN ATM PON spec has been adopted as ITU-T Recommendation G.983. Despite the efforts of FSAN, Goldman doesn't expect FTTH to be economically feasible until 2001. "The key issue still holding people back," says Ross, "is the powering issue." The optical receiver/transmitter, says Spears, "needs to be powered locally, with or without battery backup," depending on the service set.
At some point, Ross believes, subscribers will have to take responsibility for powering locally, just the way homeowners today maintain their smoke alarms. He foresees a signaling scheme built into local power units that, like smoke alarms, alerts subs that power is low.
The recent initiatives by BellSouth reflect a fairly long commitment by the company to fiber technologies. Goldman says that BellSouth first conducted FTTH trials as long as 12 years ago in two subdivisions in central Florida. Back then, the equipment rack necessary could fill a customer's garage. Now, it's the size of a modem.
Clearly, the economics of deploying fiber are a factor in the company's deployments, says Goldman. While the price of the optical electronics has fallen, the life cycle and maintenance costs of fiber are more attractive. Goldman explains that with fiber, adding electronics such as dense wavelength division multiplexing to the network can increase capacity without digging up streets and laying additional fiber. This becomes important in fast-growing areas.
Despite occasional fiber cuts, "fiber tends to be maintenance-free," says Goldman, adding that copper can be subject to interference, noise and other problems. Thus the life cycle maintenance costs of copper are factored into the business case for laying fiber. Often, these cost savings can help offset the initial cost of installing fiber, especially in areas where the costs of maintaining copper are high. Exactly where BellSouth decides to introduce fiber is also a function of location, as some areas are more expensive than others. Ultimately, BellSouth sees its passive optical network strategy as a way to "future-proof" its network.
Obviously, only time will tell if these operators have done all they can do to anticipate the bandwidth needs of their customers for the next few years, but clearly, the race is on.