Despite reports to the contrary and the documented financial struggles of many Data Local Exchange Carriers (DLECs), the Digital Subscriber Line (DSL) market still remains hot. As many DSL advocates on both the residential and business side of the fence can attest, DSL is a great alternative to the slow modem wasteland of dial-up, the per-minute usage charges of ISDN, and the high costs of T-1 or even fractional T-1's. However, the core problem with Central Office (CO)-based DSL still remains the distance limitation of the local loop. Unless a customer is within a specific distance from the CO, that being around 18,000 feet, the customer is pretty much out of luck if it wants service. Even if a customer can get service, the quality of service it receives is pretty much dictated by how far it is from the CO.

In addition to being harnessed by distance limitations, many CO-based DSL providers are also struggling with the market's high entry and customer acquisition costs. Not only do they have to register as Competitive Local Exchange Carrier (CLECs) but they must also collocate at the CO, which may require huge capital investment. These factors have driven many providers to seek out new ways to deliver robust DSL services to customers outside of the CO. Up until recently, the options have been pretty limited.

CO-Based DSL: Not Making The Grade

Why the big rush for providers to step beyond the CO? Well, even if the customer is fortunate enough to fall within the required distance of the CO, chances are it is not going to get the speed or reliability it signed up for. The reason for this is due in part because most CO-based DSL providers utilize Asymmetric DSL (ADSL). ADSL offers faster speeds downstream (from the Internet to the customer PC) than upstream (from the customer back to the Internet). ADSL was designed primarily for residential and home office type customers and was not designed to meet the higher bandwidth needs of the larger small and medium sized business community. Thus, even if the customer signed up for 384 Kbps, chances are it will not receive that high level of bandwidth on a consistent basis, unless its business is located directly outside of the CO.

Also, many CO-based providers are unable to take advantage of future DSL technologies promising much higher access speeds because of the loop limitations. This includes Very High Bit Rate DSL (VDSL), which promises to deliver up to 13 Mbps bandwidth speed, but is limited to loops of up to 4,000 feet. Another issue that compounds the distance limitation problem is loop quality. Such things as electrical noise, as well as broken drainage systems flooding binder groups can affect loop quality. Thus, the greater the distance from the CO, the greater the opportunity for loop deterioration.

The loop reach limitations of CO-based DSL have spawned various flavors of DSL with the goal of improving the level and reliability of service. Some examples of these DSL flavors include G.lite and IDSL. G.Lite provides customers with an analog-to-digital modem ensuring that the transmission from the phone company is digital rather than the analog transmission of "plain old telephone service." However, while G.Lite does offer higher speeds at 18,000 feet than ADSL, it still has the same reliability and distance limitation issues. When IDSL digital data is transmitted at 128 Kbps on a regular copper telephone line from a user to a destination using a digital transmission, it bypasses the telephone company's CO equipment that handles analog signals. While this may be a marked improvement, the maximum speed of an IDSL circuit is only 128Kbps, really only the first step-up from dial-up. Businesses that use IDSL and require higher bandwidth speeds will have to rely on providers to bond together multiple circuits in order to deliver higher bandwidth speeds.

Understanding the limitations of ADSL, and its not-so-great alternatives G.lite and IDSL, are there any other options out there? Well yes, sort of. Symmetric DSL (SDSL) delivers up to 2.3 Mbps and offers the same speed in both directions, serving as the equivalent to business-grade DSL. Even so, there is a significant drawback to SDSL in that it required short loops to deliver the higher levels of bandwidth speed. Since the majority of customers are beyond 15,000 feet, SDSL is not the end-all solution to the distance limitation problem.

In addition to the distance limitation and loop quality issues, many providers have found significant entry barriers into the DSL market. Becoming a CO-based DSL provider is a long, expensive, and arduous process. Until recently, a DLEC's route to becoming a DSL provider first involved registering as a CLEC. This is because only CLECs are allowed to install DSL equipment in collocation cages at the Incumbent Local Exchange Carrier's (ILEC's) CO. Once the provider has registered as a CLEC, it must then begin the process of negotiating with the ILEC for collocation space in the CO. The next step involves negotiations with the ILEC to obtain access rights to the local loops. Collocation is long and expensive process and may take anywhere from six to eight months before a DSL provider can begin offering services to customers.

So how do providers solve the loop distance, quality, and collocation issues to deliver fast and reliable service profitability? One popular alternative is for DSL providers to move away from the CO, build their own infrastructure, and deliver the DSL technology closer to customer. Bringing the DSL technology closer to the customer eliminates many of loop limitation problems and provides much shorter runs from the customer premise to the where the DSL equipment is installed.

Lets examine a few different approaches providers have taken to solve the last mile access problems.

Building-Centric or BLEC Approach

One of the simplest and most cost effective approaches is the building-centric or 'BLEC' model. In this approach, providers essentially take the DSL technology traditionally found at the CO, minimize it in the form of mini-DSLAM and place it in the telephone closet of Multi Tenant Units (MTUs) or Multi Dwelling Units (MDUs). The provider then plugs the DSLAM into the building's existing copper infrastructure. The in-building runs are much shorter than CO runs, in that the distance from the basement to the top floor of a building will be at most 1,000 feet. This deployment strategy requires only a few flavors of DSL, since over 95 percent of the customers are within 3,000 to 4,000 feet from the basement of the building. This allows the DSL provider to offer the highest speeds to the tenants of the building.

Also, this model enables providers to utilize SDSL (delivering speeds up to 2.3 Mbps) and the model guarantees customers consistently high speeds because of the shorter loops. The quality of the loops is not compromised and remains stable and consistent throughout. Why? It does not extend beyond the building. The key benefits of this deployment strategy include:

* Requires less initial upfront investment
By utilizing the mini-DSLAM equipment, providers are able to deploy service in a building for $10,000 to $12,000 on average. This is a fraction of what a CO deployment costs which often requires an upfront investment of about $50,000 to $100,000, in addition to expensive monthly collocation rental costs.

* Less dependence on the RBOC
CO-based providers are at the mercy of an ILEC, which can take three to four weeks to qualify a loop to the customer. Time to provision may take even longer. These delays often damage the credibility of providers and take away much needed revenue. Remember the RBOC is often competing for the same customers as the DLEC, so it will try to put up as many barriers as possible to limit entry into the market. In contrast, BLECs focus specifically on providing service to MTUs and MDUs. They minimize their use of the ILEC infrastructure by installing equipment in every building they serve. As soon as the service provider strikes a deal with property management, the provisioning process begins.

* Faster service deployment
Since BLECs do not have to depend on the RBOCs for the loops, they can deploy service in a matter of days in the buildings as opposed to months from the CO. The only area where the BLEC is dependent on the RBOC is the provisioning of the high speed Internet link, usually a T-1, to the building. BLECs do, however, have the option of provisioning the T-1 through a CLEC rather than the RBOC if it is proven to be more cost-effective.

* Lower cost of winning customers
CO-based providers have a broad potential customer base to market their services toward and therefore tend to advertise on TV, radio, and in local and trade publications. This often adds up to huge marketing costs. On the other hand, BLECs have the unique ability to focus intensely on a smaller set of customers in order to achieve higher levels of penetration and increased Average Revenue Per User (ARPU.) This strategy requires much smaller fixed investments per building that are easy to reallocate if necessary. However, BLECs require a much higher market penetration rate in order to become profitable.

* Less competition from other service providers
Most CO-based providers must share co-location cages in the CO with other DLECs. Each of these DLECs ends up competing for the same customers with the ILEC. Since there is less space in the basement of buildings, BLECs encounter fewer competitors in any given building and for the majority of time, they are the first service provider to deploy DSL in the building.

* Lower break-even point
CO-based DSL requires 300-400 business customers per CO to be profitable in three years. In contrast, a BLEC, on average, only requires an acquisition rate of six customers per building to be profitable within one year due to the low deployment cost.

* No CLEC registration or interconnection agreement
CO-based providers have to file for CLEC status and negotiate interconnection agreements separately in each RBOC territory causing a huge delay and significant legal fees. Because the BLEC only deploys its equipment in the buildings and does not interconnect or rent space within the CO, there is no need to file to become a CLEC. This frees up significant funds that can be used in deploying service.

In addition to commercial office properties, the BLEC approach is also an attractive strategy when targeting MDUs like apartment complexes and campus environments. In MDU environments, the majority of the copper wires go back to a central location. Thus, a DSL network can be quickly established by simply installing a DSLAM in in that central location.

The BLEC deployment strategy eliminates much of the significant loop distance, collocation cost, and regulatory and provisioning hurdles faced by CO-based providers. However, BLECs are limited to marketing their services in buildings that they have provisioned. In addition, they often have to secure revenue sharing or rental agreements with building owners or managers to obtain access and place their equipment in the building.

The Next Door Approach

Rather than collocate facilities in a cage in the CO, some providers have chosen to install facilities in buildings next to the CO. The provider simply installs a DSL concentrator or DSLAM in the building and picks up loops outside of the CO. This deployment model eliminates the need for a provider to rent cage space and collocate with the ILEC in the CO. Thus the provider does not have to register as a CLEC, eliminating many of the regulatory hurdles and legal costs it may have otherwise faced. This strategy also reduces time to market for the provider and quickly deploys points of presence (PoPs) in targeted service areas. Once the adjacent building has been chosen, the provider then reserves copper pairs from the CO to the building where the pairs terminate on a cross connect. The cost of these copper pairs can range from $5 to $20 per loop depending on the market.

The Next Door approach is a viable solution to the collocation problems CO-based providers face, in that many experts predict that COs will run out of cage space in the next year or two. The approach is also very cost and time-effective as it allows providers to quickly deploy service. While collocation in the CO typically takes months, local loops can be provisioned almost immediately. This approach gives providers an alternative to becoming a CLEC and allows them to begin offering services to customers right away. As collocation space begins to run out, this strategy should become a more and more popular alternative for service providers.

Passive Optical Networks (PONs)

Passive Optical Networking (PON) is another networking strategy aimed at solving the problem of last-mile access. The intention is to bring the same level of bandwidth that exists at the center of the network to the end users. A PON is basically created when the provider installs fiber that runs from the CO or POP to a cluster of end-users. The data traffic hits a splitter that directs each customer's traffic to a link that extends from the splitter point to the customer premise. Often this network configuration is described as a "tree" because of the fiber coming from the CO looks like a trunk and branches out into multiple lines from the splitter. PONs are passive in that the fiber optical transmission has no power requirements or active electronics once the signal travels through the network. Power is only utilized at the CO and customer premise.

The main benefit of a PON is its ability to extend fiber closer to the end-user. With PONs, only short runs of fiber or DSL are required to connect customers to the PON trunk. DSL and PONs have a complimentary relationship. By moving the PONs closer to the customer, the DSL links can run from the splitter point rather than the CO. This greatly reduces the distance limitations often found with DSL service. There is also a great cost efficiency derived from utilizing DSL in the PONs network. Using DSL links from the fiber optic PON means that carriers do not have to deploy fiber all the way to the customer premise, instead handing off the traffic to the DSL line, which completes the transmission to the customer. This use of shorter DSL runs allows providers to deliver more reliable and faster service.

Deploying PONs as an alternative to CO-Based DSL has significant merit because it improves the quality and speed of service by adding fiber to the network rather than relying entirely on the copper-based local loop from the CO to the customer premise. The shorter runs also gives providers the option of using SDSL instead of ADSL because loop runs from the PON to the customer may fall within the distance required for higher speed SDSL service.

Given the fact that fiber tends to be less available and more expensive for small business locations than large business locations, the PON allows for providers to bring broadband services to small businesses. By using fiber, small business customers are able to gain access to optical networks from which they would normally be excluded. A PON makes it affordable for smaller businesses to get fiber access by sharing the fiber bandwidth with multiple subscribers.

PONs are also an excellent deployment strategy for service providers going after customers in a more dispersed environments, such as office parks. However, it should be noted that unless a provider already has fiber in the ground, PONs may not be a smart deployment strategy as digging up the ground to run fiber tends not to be an effective business case.

There is a widely held belief in the industry that PONs will not scale adequately to meet the capacity demands of future applications. Part of the reasoning behind this belief is that because of distance limitations and the fundamental belief that fiber is much more superior to copper, DSL cannot be counted on to deliver the high margin services of the future. However, while PONs may not be the permanent fix to the access problem many analysts have been looking for, they serve as an ample substitute until fiber can be directly deployed to the customer premise

Whether it be through BLECs, PONs or even fixed wireless solutions, it is becoming more and more evident that the last-mile access dilemma will inevitably be solved. CLECs, DLECs and BLECs are deploying new technologies and networks designed to lessen their dependence on the ILEC and allow them to deploy service at a much quicker and cost-effective rate. As more competitors are able to deploy their own infrastructures, whether it be through the utilization of fiber optic strands, in-building networks, or their own POPs, providers will soon be able to deploy reliable high speed DSL services to the majority of small and medium sized businesses.

About the Author

Chris Zanardi is Communications Director at Wired Business. Wired Business is a building-centric provider of high speed Internet access and data communications services to small and medium sized businesses in multi-tenant commercial office properties. The author may be reached with questions or comments via email at czanardi@wiredbusiness.com