Fiber Performance

Posted: 12/15/1999

Fiber Performance
Passive Optical Networking Takes the Stage
By Anthony Zona

Despite a heavy courtship by venture capitalists and Wall Street over the past year, competitive local exchange carriers (CLECs) are still held to the most fundamental principle of business--performance matters. With this in mind, future CLEC success will be derived from investing in assets that steadily increase over time, sustain low ongoing costs, are highly flexible and have unchallenged credibility. There is only one such asset in all of networking--fiber.

While copper still is the dominant access technology, it is not a strategic investment for CLECs. Copper has limited bandwidth, and, therefore, constrains the types of services a CLEC can sell to business customers.

This logic does not demand a quantum leap of faith. Today's carriers are deploying fiber rather than putting copper in the ground. Even if a CLEC is not facilities-based, today's business and regulatory climate provides a host of opportunities to wholesale fiber from the incumbent carriers, fiber backbone builders, utilities, municipalities, cable TV operators and national carriers. Fiber is an enduring commodity that, while measured in cents per foot, can transport traffic worth thousands of dollars per month.

According to a recent study by Vertical Systems Group, Dedham, Mass., of the 235, 000 U.S. businesses with more than 75 employees, 76 percent are located within one mile of an available source of fiber. However, other studies state that only 3 percent to 5 percent of all businesses currently have a fiber interface. Fiber networking architectures, and the electrical service devices that create those architectures, are to blame.

The sad reality is that most fiber networking architectures seem to be designed more to consume capital than to generate profits. This is hardly surprising given the fact that fiber-provisioning principles were developed by regulated monopolies with little concern for initial costs, having been guaranteed a return on their investment. In a competitive market, however, the business principles of the past monopolies no longer hold. As such, should their technical principles hold?

In a competitive market, facing an incumbent with a century of service experience, CLECs realize that optimizing revenue per customer is critical. Fiber optics may well be the profit key for CLECs. Fiber can support services with bandwidth ranges from megabits to many gigabits. Fiber is a loop technology, a metro technology and a backbone technology. Its deployment cost over new rights of way is little more than copper, but its revenue potential in all its applications is staggering.

Fiber alone is only an asset in the making. To earn revenue from it, the fiber must be terminated with an optoelectronic device that couples the fiber to the consumer, or to other elements in the service provider network. This process, called "lighting the glass" by some, is usually far more expensive than the fiber itself.

The current rage in optical technology, dense wavelength-division multiplexing (DWDM), is sometimes presented as a solution to the cost-effectiveness problems of a synchronous optical network (SONET). However, most DWDM systems are simply point-to-point, capacity-multiplying devices, and beyond bandwidth capacity, offer no benefit over SONET. A few DWDM systems that provide for optical networking, meaning the ability to move traffic from fiber to fiber at the wavelength level, are beginning to emerge.

Current fiber provisioning is limited to applications that justify a very high cost. A typical user of a fiber connection today pays thousands of dollars a month for access alone. This level of cost cannot be expected to support a broad market for fiber services, yet a broad market for services is crucial to profit growth for CLECs.

To aggressively attack a broad range of business customers, CLECs need a method of deploying a completely connected fiber footprint that covers their entire territory, without being forced to supply expensive electrical devices for each fiber junction. Rather than simply pursue technologies that increase only backbone network capacity, CLECs must augment their strategies to include technologies that enable them to sell high-bandwidth, value-added services to business customers of all sizes.

Passive optical networking (PON) does that. However, while PON technology has existed for several years, its application to date has been limited mainly to broadcast applications for residential cable television. Recent advances in the technology have made PON a more compelling solution for competitive service providers. With PON, the entire access network fiber plant can be deployed with little carrying cost, and activated selectively when customers require service. This revolutionary development in optical networking will fundamentally change how competitive service providers build networks and grow market share in the business-customer segment.

Fanning Out

With optical access networking, the service provider deploys fiber media using the PON approach with passive splitters/couplers linking each fiber strand. In a typical tree structure, the trunk is the link to the service provider's main point of presence (PoP). This fiber tree can be made up of provider-installed fiber, or dark fiber, wholesaled from another provider, and the entire fiber network can be deployed at one time, or portions can be added as conditions dictate. While the network is in this initial state, it contains no electronics at all, and there is no equipment capital cost or ongoing equipment support cost to bear.

Presumably, the carrier would, at this same time, be developing a set of prospects and making sales calls. Because the fiber tree network can fan out to a large number of prospect locations, sales activity can be spread beyond a few selected buildings, as would be the case if active SONET rings were deployed.

When the first customer is added, the carrier installs a central office (CO)/head-end services aggregation device, or optical access switch (OAS), at the trunk end of the optical tree that serves the customer's location or locations. At the tip of each service branch supporting a customer site, an intelligent optical terminal (IOT) is installed. The two-way communication between the OAS and IOTs are carried via an innovative protocol known as dynamic wavelength slicing.

Dynamic wavelength slicing allows individual wavelengths to be sliced, or shared, among many end users. Just as DWDM has expanded the capacity of a given fiber, dynamic wavelength slicing expands the number of termination points that can be served over a given wavelength. Each wavelength used in the fiber tree originates in the OAS and terminates in some number of IOTs. The wavelength is sliced dynamically to permit each IOT to have full-duplex communications supporting a variety of service interfaces. This allows bandwidth to be delivered in a wide range of quantities to individual IOTs, which supports the bandwidth needs of more typical network users. Because multiple users share a given PON rather than each being allocated an entire wavelength, the total capacity of the fiber is utilized more efficiently. In addition to providing customers with the amount of bandwidth they need today, dynamic wavelength slicing also enables businesses to request additional bandwidth to address situational or seasonal fluctuations in consumption. This gives service providers a cost-effective way of deploying a wide range of service tiers to customers, tailor made to the needs of the individual customer.

Intelligent optical terminals at the ends of the tree's branches can be used to support individual customers with relatively high demands for bandwidth, such as corporate headquarters or data centers. They can also support shared-tenant applications in buildings or office parks.

A complex fiber network made up of passively coupled strands can be deployed to cover a huge service geography by leveraging embedded fiber assets and/or rights of way, because the only cost is the media itself. As customers are added to the network, the service provider selectively adds IOTs and OASs to provide service to additional customers.

Broad Appeal

PON has long been recognized as the least expensive way to deploy an access fiber infrastructure. With low-cost fiber assets to link key customers to a service network, the CLEC can open a wide variety of profit sources by introducing both CLEC-based services and cooperating with regional and national carriers to offer a full range of services.

Just as optical access networking supports multiple service strategies for CLECs, it also supports multiple marketing strategies. CLEC success is heavily dependent on choosing the best way to market services and develop prospects into customers. The nature of the prospects in the territory, the type and quality of the salespeople available and the nature of the competition may dictate the CLEC's approach. It should never be dictated by the network technology, yet that may well be the case with traditional optical network architectures. With these traditional approaches, the CLEC may be forced to sell based on the initial equipment positioning to contain costs. This may also contain opportunity.

All the alternative architectures for CLEC service networks are designed to maximize the revenue opportunity through the concept of the opportunity cluster; the grouping of prospects with a combined business that can justify the CLEC's entry into the provider market. Opportunity clusters are most often created by forces outside the CLEC's direct control--groupings of customers in a common vertical market, a specific problem with an incumbent's performance in a given area or simply rapid growth in a community or region. Whatever the factors that create the cluster, optical access networking can allow the CLEC to develop those opportunities with the lowest possible startup cost and the best possible ongoing return on capital.

In the wake of telecom reform, and in a booming economy, CLECs have become sought after by capital sources. For some, this heady period of credibility with Wall Street may have created a false sense of security. A glance at the regulatory news shows the competition in the local exchange market has barely begun, and some analysts predict as many as 60 percent of all current CLECs will not survive the next five years.

For all CLECs, the key to success is the digital exploitation of key service customers. For many, because of the relative ease of gaining access to copper loop, that exploitation has been planned on the basis of a generic digital subscriber line (xDSL). However, xDSL offers only a few megabits of digital capacity, less than medium-sized to large sites may already consume. What's more, the use of fiber remotes may completely invalidate xDSL services to many key opportunity areas. Nonetheless, in areas where digital loop carrier (DLC) has not been deployed, xDSL is an excellent solution for at-home users. As such, optical access networking represents a way to cost-effectively exploit fiber to mitigate the service delivery shortcomings of the legacy copper infrastructure and quench the pent up demand of business locations.

Optical access networking is the only fiber networking architecture to utilize the PON technology that shows such promise even in low-end residential applications. With PON, optical deployment costs plummet and CLECs are able to prepare their entire optical plant for customer service while lighting only that portion of the plant actually serving customers. This combination reduces early cash demands and increases the ability of CLEC sales forces to turn on new customers by eliminating geographic restrictions caused by early equipment placement.

Anthony Zona is founder, president and CEO of Quantum Bridge Communications Inc., North Andover, Mass. He can be reached at (978) 688-9100.

Share this article: Email, Slashdot, Digg, Del.icio.us, Yahoo!MyWeb, Windows Live Favorites, Furl
Add this article feed to: RSS, My Yahoo, Newsgator, Bloglines