May 2001

BY DAVID KAMM

Those familiar with Digital Subscriber Line (DSL) technology may also recognize the term "in-building DSL," meaning the deployment of DSL inside a building using in-building copper pairs, versus using DSL as the last-mile technology connecting a building to a service provider. In-building DSL is an appropriate and useful technology for hundreds of thousands of commercial multi-tenant units (MTUs) in the U.S. and other regions of the world, especially where these buildings (and their tenants) can not be reached with local-loop DSL services.

Another term you're likely familiar with is Voice over Broadband (VoB), which encompasses a number of methods for delivering packet telephony services over high-speed access connections. Voice over DSL (VoDSL) is a specific example of VoB, and it is gaining widespread acceptance as a viable, cost-effective approach for delivering multiple voice lines, as well as high-speed data services, over a single twisted copper pair. According to TeleChoice, more than 20 service providers had announced commercial VoDSL services in the U.S. and Canada as of Q2 2000. Deployments of VoDSL are expected to ramp up dramatically as new providers conclude field trials and begin to aggressively market bundled voice and data services to end users. Key markets for VoDSL (and VoB in general) include millions of small to midsize businesses (SMBs) as well as residential users with multiple phone lines and computers in the home or home office.

Adding Voice To A DSL Network
Adding packet voice to a previously data-only DSL network places new demands on the access network and the in-building equipment. Whereas data-only services can generally get by with best effort Quality of Service (QoS) characteristics, packet voice services must have low, predictable delay (latency) in order to ensure toll-quality voice.

Another factor -- amplified by the clustered nature of MTU-based subscribers -- is that not all MTUs can get DSL services today. Many commercial buildings are simply too far from a DSL-capable Central Office (CO) to receive services. In addition to the distance-related limitations of DSL, many relatively new (less than 10 years old) commercial properties are served off of Remote Terminals (RTs), rather than directly from a central office. This adds to the DSL availability problem. Many RT-served buildings are actually close to a CO, but nevertheless are locked out of DSL service unless an RT-based or in-building DSL solution is available to them.

The problem of DSL availability in the local loop is greatly exacerbated once we start thinking about very high speed, shorter-reach DSL variants like VDSL. In fact, depending on the density of MTUs around any given central office, the number of buildings that fall outside the DSL service range can grow exponentially as the service radius is reduced from SDSL range to VDSL range. Unfortunately, this means the resultant lost revenue opportunity for ICPs grows exponentially as well.

An in-building DSL approach can bypass this problem completely. Even a "long" in-building copper loop is unlikely to exceed 1,000 feet or so, making in-building VDSL speeds up to about 50 Mbps a reality. Of course, the DSL Customer Premise Equipment (CPE) must be equipped with VDSL interfaces, and the in-building DSL concentrator must have the architecture and processing power to handle multiple VDSL feeds at full line rate.

Data-Only Services Aren't Enough
Many Integrated Communications Providers (ICPs)  and most industry analysts have come to the conclusion that data-only services in the MTU are not enough to generate long-term profitability and subscriber loyalty. A data-only approach also makes it harder to recover initial investments and recurring costs, since it requires more subscribers for the ICP to reach a break-even point. The overwhelming revenue opportunity for MTU-focused ICPs and an attractive path to profitability with broadband access is to offer packet voice services in addition to data services -- true multi-service access for tenants. VoDSL is expected to be a highly profitable service offering for ICPs, with excellent return on investment and net present value characteristics over a wide range of pricing structures. Similarly, end users stand to benefit from reduced bundled pricing, a single point of customer contact, and simplified billing.

For this article, let's concentrate on delivering VoB services to MTUs via in-building DSL where DSL as a local loop option isn't available. Some analyses indicate that when both fixed and monthly recurring costs are factored in, in-building concentration can be more cost effective than wiring each business/subscriber directly back to a CO-based DSLAM. This is due in part to the fact that with in-building concentration, fewer uplinks are needed and oversubscription is an option. Service providers can also leverage an in-building concentration strategy to avoid the recurring delays associated with provisioning individual subscriber lines from another telco, saving time and money. By utilizing a platform called an Integrated Concentration Device (ICD), the ICP has the option of wiring phones, key systems and PBXs directly to the ICD in the basement or wiring center, avoiding the purchase of dedicated IADs for each subscriber.

New Options For Deploying MTU-Based VoDSL
An ICD configuration (depicted below) is a new alternative for MTU-based VoDSL. It integrates subscriber voice CPE such as phones and key systems directly into the concentration platform, allowing the ICP greater freedom to offer different subscriber services with one platform.

This approach recognizes and leverages the fact that the MTU-based subscribers are co-located with each other, which provides much greater deployment flexibility. 

Conclusions
The opportunity for packet voice services in MTUs is great, and the revenue potential very large, especially when compared to a data-only MTU services strategy. Service providers have a number of choices for delivering VoDSL in MTU environments, including a new approach that incorporates both DSL and voice customer premise equipment in a single concentrator.

An in-building DSL strategy can be an excellent alternative to local loop DSL, even where subscribers are close enough to be served from CO-based DSLAMs. In-building DSL solves the well-known distance and reach problems associated with VoDSL, and is attractive for many more reasons as well.

Adding DSL-based packet voice services in MTUs presents the first real QoS test for in-building DSL concentrators. Concentration products designed primarily for MTU-based data services will be the weakest link in the multi-service access chain. Only platforms based on true multi-service architectures will succeed as the foundations for toll-quality voice, broadcast-quality video, and other QoS-intensive, highly profitable services in the MTU.

David Kamm is the director of marketing for Avail Networks, Inc. Avail Networks designs, develops, and markets network access solutions worldwide for next-generation convergence applications such as integrated data, voice, and video over single broadband connections. Avail's flagship products -- the Frontera family of multi-service access platforms -- enable service providers to deliver integrated broadband communications services to end-user sites in commercial office buildings, office parks, and single-business locations.

For additional information on standards-based voice over broadband topics, visit www.openvob.org.