Fiber trends look positive for metropolitan/access applications
Bandwidth demands and competitive pressures will drive fiber deeper into access networks between now and 2003.
Jeff D. Montgomery,
Stephen Montgomery, and
Peter T. Jewett
ElectroniCast Corp.
President Clinton signed into law a major overhaul of the U.S. communications industry in February 1996. Subsequent challenges filed in court may delay the final realization of the law for a few years. It appears certain, however, that as we move into 1999, any company that wishes will eventually be able to enter any segment of the communications service industry, including land-line and wireless voice, data, and image communication; video to the home; virtual local area network services; and voice over the Internet/voice over Internet protocol (IP). Here are a few examples:
Interexchange carriers (IXCs) may expand into the local access and transport area to provide business and residential voice, data, and video service.
Incumbent local-exchange carriers (ILECs) may expand into interexchange transport service, as well as into residential video service--including combined video and telephone service in the nominal territory of competitive LECs (CLECs).
Cable-TV system operators may expand into voice and data transport, via land-line and wireless, to residential, business, and interexchange.
Competitive access providers (CAPs) and Internet service providers (ISPs) previously limited to linkage between business subscribers and interexchange points of presence (PoPs) may now provide any service they wish.
Anyone may enter this competition as a startup, leasing central-office switching capacity from LECs and transport capacity from any of several suppliers--no owned hardware necessary. Even utility companies are exploring and testing opportunities along their rights-of-way.
There are now roughly 25 to 30 major competitors in this newly defined North American communications-services market, counting telephone companies, IXCs, major wireless networks, cable-TV, and multiple-service operators. These areas account for well over 90% of communications-services revenues. Several hundred smaller suppliers provide most of the balance. At the same time, the formation of sales/management-oriented organizations that specialize in "freight" packaging/forwarding for the information highway (without owning the highway, tollgates, or trucks) will likely attract major capital investment.
Not surprisingly, considering the large number of players, market consolidation has been proceeding for years, affecting independent LECs, CAPs, and cable-TV firms. In the new environment, this trend will accelerate. More multibillion-dollar mergers, comparable to AT&T/McCaw, WorldCom/Brooks Fiber Properties, WorldCom/MFS, AT&T/TCI, and AT&T/TCG lie ahead.
Fiber "toll roads" will expand
In the battle for continental (and global) market share, the major selling points will be fast connection, flexible access to hundreds of megabits of bandwidth, and high reliability. Achieving these attributes will require fiber extension (versus coaxial or twisted-pair cable) to subscribers, more redundant loops, and expansion of feeder, interoffice, and interexchange capacity via more fiber, higher data rates, and DWDM.
Concurrently, there is a growing demand for increased reliability in telecommunications networks. One of the largest problems with fiber-optic networks is the possibility of a backhoe cut. This type of break takes out a whole cable, not just a single fiber. Thus, to increase reliability, route diversity is needed, so if there is a cut in one cable the traffic can be instantly routed over another route to the destination.
In the context of the increased deployment of fiber that these two trends will spark, it is worth noting that a significant share of broadband communication now flows over fiber capacity leased from public utilities. Since they have rights-of-way and cable installation crews, utilities can provide and expand this capacity, continent-wide, at the lowest cost and risk. These factors will be significant in North American bandwidth capacity expansion over the next decade.
Competition fuels new networks
There will be continuing growth in data-communications traffic, and this traffic will increasingly require higher speeds. Data-communications traffic was more than 50% of the total traffic in 1998 in the public networks. The balance was, of course, voice traffic. On private networks, data-communications traffic is already more than 60% of the total. In terms of new services, IXCs are offering frame relay and are working with the LECs to offer other services. These new services boast scaleable transmission rates, so as demand goes up, the carriers can increase the rate. All types of communications service providers are moving aggressively to explore the benefits and opportunities provided by IP technology.
This rapid growth in network bandwidth will be augmented by increasing duplication (overlay) of broadband communications networks. In North America, as LECs, IXCs, cable-TV operators, and other service providers compete to provide broadband services, it will often be more feasible for a new competitor to install a new network to the customer, rather than negotiate for access to an existing network.
Two factors will influence this trend. First, there will, in general, be substantial delays in negotiations before a competitor can gain access to existing networks. Second, the existing network often will not meet the needs of the new competitor. While this increased competition will be at its highest level in North America, such duplication also should increase in Europe, Japan, and other world regions over the next decade.
SONET affects metropolitan/access
One metropolitan/access trend is the continuing deployment of fiber-optic SONET self-healing rings. Such rings are difficult to categorize since they provide both interoffice and feeder functions. Regardless, optical fiber and SONET are working their way down into distribution networks. Interoffice links (central office-to-central office) comprise fiber for all new builds. The feeder portions of LEC networks are on the order of 50% to 60% fiber now; thus, there will be room for continued consumption in that area.
In the residential distribution category, there is almost no fiber now. Therefore, there is potential for large growth in the number of links here as well. However, these links will be relatively short in length, and thus represent only a modest market potential as far as total miles of fiber is concerned. Optical fiber to business users, meanwhile, often utilizes only the distribution link (eliminating the feeder). Therefore, in major metropolitan areas that have high-density subscriber populations, optical fiber already passes and/or drops at commercial buildings and campuses in many cases.
The SONET-based self-healing rings not only can provide route diversity to business customers that require high reliability but can be used by the ILECs to offer new services with minimal capital investment. For example, one of the services offered to business clients starting in 1993 was switched multimegabit data service (SMDS). With central offices tied together by fiber rings, the SMDS switch is in one central office that can serve the whole local access and transport area (LATA). As SMDS traffic increases, additional central offices are outfitted with SMDS switches as the need arises.
In 1996, ILECs also began to offer ATM switching as a service. As the traffic increases and as operating system development allows, this service will continue to be integrated into the network. However, IP over SONET is gaining a lot of attention, since new carriers typically choose to transport their traffic in native IP format rather than in ATM format. At the same time, business customers are driving the deliverance of IP from the CLECs. The CLECs require products that can be installed easily and rapidly. CLEC customers are typically commercial clients and businesses that already have local-area/premises data networks established and need to extend their networks over the metropolitan-area and wide-area networks.
ILECs and other carriers will offer personal communication services (PCS) by the first years of the next century as well. These networks will have enough capacity to provide wireless service to many subscribers. However, PCS will not have much of an impact on the consumption of fiber-optic components in telecommunications, due to the narrowband nature of each PCS connection. Much of the fiber that is already in place could pass traffic among PCS wireless nodes.
The advent of PCS will cause little net increase in traffic; most links will mainly carry traffic that would otherwise be on a wireline network. Thus, ILECs are working on advanced intelligent network (AIN) plans and implementation. These activities require very little increase in bandwidth to implement and requires mainly signaling protocols, computers, and on-line databases. However, PCS will demand the use of additional fiber-optic base stations and nodes, which will lead to an increased need to test at the nodes.
North America leads consumption
Our research indicates that the North American (United States, Mexico, and Canada) consumption and deployment of fiber-optic cable in metropolitan/access networks held a relative market share of 50%, or an estimated $1.87 billion, of the global market in 1998 (see Table 1). The North American share will decrease to 44% in 2003; however, the consumption value will climb to more than $3.7 billion. In 1998, the United States represented about 89% of the North American demand; the U.S. share will decrease to 84%, or $3.1 billion, by 2003.
North America also led consumption of optical-fiber amplifiers used in metropolitan/access networks. These were basically all erbium-doped fiber amplifiers (EDFAs) in 1998, and EDFAs will dominate optical-fiber amplifier metropolitan/access use over the 1998-to-2003 period (see Table 2). These amplifiers flow into telecommunications applications mainly through two channels:
amplifier modules built into network products, such as digital crossconnect switches and SONET multiplexer terminals, either purchased by OEMs or captive production within vertically integrated OEMs
DWDM link systems, which typically are optical-amplifier-augmented fiber links, including transponders/transmitters, receivers, and WDM multiplexers, sold mainly to end users such as IXCs and LECs.
ILECs to dominate
During 1991 to 1998, cable-TV multiple-service operators accounted for most of the North American distribution fiber deployment. In 1999 and beyond, the ILECs will become the largest consumers of fiber-optic cable for distribution applications, followed by cable TV. The next largest consumer of distribution fiber-optic cable will be the U.S. independent telephone companies, IXCs, and CLECs/CAPs. In distant third and fourth places will be the Canadian phone companies and TelMex. Canadian and Mexican deployment of distribution singlemode fiber cable will rise from a negligible 1997 level (trials). Although the growth in the latter part of the 1999-to-2008 period will be rapid, the quantities will still be relatively small.
Within the context of the U.S. marketplace, alternative local carrier (ALC) operators (also referred to as CAPs) have become strong competitors for local-exchange companies in the urban market. Their success is based on the aggressive posture bypass operators have assumed in terms of service and technology. The emphasis has been on fiber-transmission facilities since metropolitan area networks were first introduced in 1987, and bypass operators (CAPs) introduced digital hubbing and 64-kbit/sec clear-channel services sooner than the regional Bell operating companies and CLECs.
Downtown office buildings represent the primary market for ALCs. The typical downtown network is based on a redundant ring arrangement several miles in length, with 100 to more than 400 fibers per cable. At least one fiber pair is dedicated to selected customers, usually electrically allocated. Cables emanating from a central hub location are typically deployed in subway or steam tunnels or other existing conduits.
In addition to continuing to expand their fiber-optic rings within downtown areas, these companies are now beginning to expand into the suburbs by means of larger rings. The intention here is to capture the businesses, many of which have moved from downtown urban areas to outlying industrial and commercial parks.
Competition and technology drive growth
Deregulation and growing bandwidth demands have attracted a wide variety of new service providers to the communications marketplace. As these carriers attempt to establish a competitive advantage by providing a broader range of high-bandwidth services directly to their customers, they will increasingly turn to fiber as their connection medium of choice.
While this trend is particularly apparent in North America, our research shows that other regions of the world will follow this lead. u
Jeff D. Montgomery is chairman, Stephen Montgomery is president, and Peter T. Jewett is a research analyst at ElectroniCast Corp. (San Mateo, CA).