DWDM in the metro market: Not a question of why-simply of how

Long-distance companies were the first to invest in dense wavelength division multiplexing on a large scale. The financial success of long-distance DWDM vendors testifies to the technology's maturity and effectiveness in multiplying existing fiber network capacity while keeping infrastructure costs low.

Indeed, the bandwidth squeeze is the fundamental driver behind DWDM's mass deployment in the long-distance market and is at play in the next market to experience full-scale DWDM deployment-network operators in the metropolitan arena. This intensely competitive market includes incumbent local exchange carriers, competitive LECs and competitive access providers.

Gilder's law and the short-haul market The number of Internet users is predicted to top 200 million by 1999, but the Internet is just one phenomenon causing the bandwidth squeeze. On the horizon are new applications such as high-definition TV, video-on-demand, telemedicine, collaborative engineering and videoconferencing. Ultimately, all these applications and services must be delivered to the end user.

Although average bandwidth supply in long-haul segments continues to follow Gilder's law faithfully-three-fold increases every year-the story is somewhat different for the short-haul segments in the telecommunications marketplace. In the metro environment, the bandwidth supply is not keeping pace with Gilder's Law, especially closer to the end user.

The burden of bringing high-bandwidth applications to the end user falls on services such as digital subscriber line (DSL), ISDN, T-1, T-3, asynchronous transfer mode and switched multimegabit data service. Within the customer facility, 100 Mb/s fast Ethernet and Gb/s Ethernet are used for interoffice traffic.

DWDM systems in the metro environment will start to blur these boundaries as new services are offered to the end user. This is the arena in which metro operators must provide service. Therefore, these operators must expand the capacity of their fiber, a vital necessity in delivering such a broad mix of applications and services.

But laying new fiber to increase capacity usually is not an option. Many metro operators lack the capital resources even to consider physically adding fiber to the network. The cost of laying new fiber in metropolitan areas often is prohibitively high and can be compounded by long litigation or regulatory delays.

The industry standard for laying new fiber in urban areas is about $80,000 per kilometer, which translates into a $24 million price tag to plow in a 300 kilometer metro ring and up to $8 million to run a basic 100 kilometer point-to-point link. Given such hefty capital requirements, it is easy to see why DWDM is the only viable option for metro operators.

Metro needs a different DWDM The choice is clear. The critical factors altering the economic logic for metro network operators is the uncertain and chaotic demand for both capacity and the mix of services offered. Moreover, faced with fierce competition intensified by deregulation, a metro operator's success hinges on its ability to respond to customer demands on short notice.

To a network already strapped for cash in the midst of this disordered environment, the choice of a DWDM system that offers a low entry cost and is sufficiently modular to accommodate pay-as-you-grow upgrades for increased capacity is paramount.

Uncertainty in the relative mix of broadband applications and services-data, voice or video over DSL; 100 Mb/s fast Ethernet; gigabit Ethernet; ATM; frame relay; or Sonet-dictates that the DWDM system of choice be channel-transparent to any data format or bit rate.

Vendors of long-haul DWDM systems would like to fill the potentially lucrative metro bandwidth void. But metro operators will find that deploying a long-haul DWDM system-even one retrofitted for metro-is inadequate and costly. Because long-haul systems are designed for transmitting a fixed and predictable range of traffic-both volume and type-their system architectures are inherently unsuitable to metro's unique needs.

In addition, design techniques used in long-haul DWDM systems to overcome impairments associated with long fiber transmissions carry a cost that makes them impractical for metro applications.

An open, scalable and flexible architecture is needed. The chosen metro system also must be totally channel-transparent and provide standardized system management that addresses the investment in legacy management systems, while providing a migration path to next generation management architectures.

System flexibility and scalability The architecture of a DWDM system determines how well it responds to the chaotic growth of bandwidth demand in the metro arena.

Typically, metropolitan centers are characterized by a concentration of business, governmental and educational institutions, each representing a user type that responds differently to new applications and service features. For example, if a new application is competitively priced, it can trigger an instant and unpredictable demand for additional bandwidth because a particular population demographic takes an interest.

Therefore, it is difficult to predict which span in a network will require extra bandwidth. In short, it is impossible to predict when and in which span a network instantly will require additional bandwidth.

To accommodate such unpredictable bandwidth demand, metro operators require DWDM systems with scalable and flexible architectures. These architectures must provide the ability to pay-as-you-grow with a low entry cost, allowing the DWDM system to scale up channels in small increments as demand dictates.

For example, a metro operator today may require only a four-channel system to expand bandwidth fourfold. However, this operator may need to expand-in a short period of time-from four to 32-channel capacity or even one channel at a time.

Expansion also must be conducted in place and while the equipment is in service. At gigabit data rates, error-free expansion is essential because a one-second interruption could affect a billion bits of information.

Transparency: Misunderstood or ignored? Among the requirements unique to metro operators, the need for channel transparency is probably the least discussed. This is because transparency is typically not required in long-haul DWDM applications.

Because long-haul operators act as local loop aggregators, they typically need to accommodate only channel feeds of 622 Mb/s or 2.5 Gb/s. As a result, they are able to meet the long-haul carrier demand using rigid and closed DWDM architectures.

Metro operators face the opposite dilemma. Not only is their bandwidth demand unpredictable and muddled, but the mix of applications and services they must provide is equally uncertain. In most cases, they must offer service for a whole range of data formats and bit rates. To be cost-effective, each channel also must be totally transparent to all data formats (Figure 1).

Transparency overcomes two cost concerns. First, the network design is simplified because it does not have to match a family of channel interface options to the customer's requirements and physically change this interface every time demand for bandwidth increases. Second, sparing costs are minimized because a transparent interface can support several types of active channels.

Another common but misguided notion is that DWDM's inherent transparency extends through all systems. In practice, most long-haul DWDM systems sacrifice transparency when additional circuitry is used to support extended distances. Because long-haul feeds basically come in two flavors, this is acceptable.

But in the metro market, transparency is more important because of the myriad types of traffic.

Challenging metro network topologies The need to support flexible network topologies is more challenging in metro networks. To meet this challenge, DWDM equipment must be able to accommodate existing and expanding infrastructures by supporting ring, add/drop and point-to-point configurations.

In cases where the fiber capacity between two congested facilities needs to be multiplied, a point-to-point solution is sufficient. However, most metro backbone networks are built using Sonet rings.

Although Sonet rings can be supported using several point-to-point configurations, such an option is not economical. In this case, the most cost-effective approach is to have all DWDM channels on the ring pass through each site while using only a portion of the channels to carry traffic in and out of each site.

The add/drop function is used to accomplish this task. With add/drop, new sites can be added easily to Sonet rings without adding congestion to the existing ring (Figure 2).

In many metro areas, several rings exist within the same geographic region. By applying add/drop to these DWDM ring networks, multiple rings can be tied together. In addition, DWDM ring networks can extend service to new customers by using add/

drop to access ring bandwidth and point-to-point to extend bandwidth to the customer. Using the basic building blocks of DWDM systems-ring, add/drop and point-to-point-almost any network topology can be supported.

Metro management Managing a metro DWDM network differs vastly from the long-haul arena. In particular, widespread deployment of DWDM systems in the metro market will require a multiple access management scheme that supports the requirements of operators and end users.

Accommodating these requirements calls for both Telecommunications Management Network (TMN) and simple network management protocol (SNMP) solutions. Because of this, DWDM equipment vendors must be proficient in both DWDM technology and traditional networking.

Two new issues also emerge.

First, because the number of metro operators dwarfs the number of long-haul operators, a considerable diversity of approaches exists in how each metro operator chooses to manage its network.

Second, metro networks are flooded with a complex matrix of need-to-know requirements.

In some situations, metro operators will choose to own or lease certain pieces of equipment to minimize costs. As a result, DWDM will mix several customer links onto a common fiber. Therefore, a metro management scheme must accommodate all those with the need to know-capacity providers, wholesalers and large end users.

Because the DWDM system lies at the heart of the network, metro network management systems must be interoperable and ultimately manage the network end-to-end.

Operators must know the status of their network to ensure optimum performance, facilitate maintenance and allocate costs. They also need to know the condition of any channels leased from other operators. In each instance, every network operator in the chain has to know that the level of service paid for is being supplied.

The challenge of a workable management system is for it to grant the information required while securing the equipment from unauthorized access.

In the corporate world and among many metro operators, Internet/intranet SNMP over TCP/IP technology is used for management functions. Still, other operators have opted for TL-1 and the upcoming Q3 or CORBA protocols.

On the other end of the spectrum, large operators worldwide have embraced TMN, a standard that defines new protocols and a hierarchy of intercommunicating management systems.

However, the drawbacks of the new system are twofold. It is far from being finished, and some carriers are ahead of others in its implementation.

Although TMN is ideal for managing operator backbones and allowing the sublease of lines, it does not address all management needs. This lack of a single standard and management platform makes shared management difficult (Figure 3).

Because metro operators and end users already have embraced TCP/IP for Internet-related applications, the development of a maintenance management network-based on TCP/IP for shared use-allows for common access to need-to-know information. Meeting the management needs for metro DWDM requires equipment vendors to provide a palette of management options that supports craft, SNMP and TMN profiles and provides multiple access using Internet technologies and protocols.

The burden is on DWDM vendors to offer systems that address the unique demands of metro network operators-systems that are scalable, transparent, flexible, have open architectures, and are equipped to support a flexible management scheme. And it is up to metro operators to demand these systems.