Copper's sweet spot

Mid-sized enterprise locations remain an extremely attractive opportunity for copper-based access technology, despite recent reports touting plans for massive deployment of fiber in the loop plant. This is so because, relative to residential or large enterprise locations, bandwidth needs at mid-sized locations are modest and existing access links are primarily T1 leased lines; about 80 percent of ILEC enterprise data communications revenue is derived from T1-based services, including T1 private lines and frame relay service. Ninety percent of these T1 lines are delivered on copper. Mid-sized enterprise locations consisting of single site locations of small firms as well as the many smaller sites of major corporations account for most of the T1 leases. Smaller locations often divide a single T1 into several channels for voice trunks with the remainder for data services. Larger locations may dedicate a T1 (PRI) for connection to a PBX with a separate T1 utilized for data services.

This heavy dependence on copper-based T1 services presents formidable challenges for enterprises and carriers. Enterprises are pretty much locked in to their existing services. No price discount is offered for adding an additional T1 line and service provisioning involves starting from scratch to find suitable copper pairs (if any), site surveys, contract negotiations and purchase of new CPE. The benefit of adding a T1 line, consequently, is usually outweighed by high procurement cost and complexity.

Carriers, for their part, lack the incentive to upgrade services. Prices for legacy T1 and frame relay services are typically higher than DSL or metro Ethernet alternatives. In addition, pricing for new T1s is well below the prices of existing contracts. Consequently, a proactive move by the carrier may result in a revenue drop or loss of the account to a competitor. The cost and operational complexity associated with T1 moves, adds and changes are additional factors weighing on carrier T1 marketing strategies. It takes many months of increased revenue to offset the cost of adding a line.

Unlike residences, mid-sized enterprise locations do not require big band-width to meet their needs. Voice requirements being met today by T1 circuits will decrease with wider adoption of VoIP and data rates of 3 Mbps to 10 Mbps provide quite a lot of headroom compared to today's typical 300 kbps to 1.5 Mbps rates.

[ILLUSTRATION OMITTED]

An ideal data rate would be 10 Mbps, which would be compatible with the very low cost and well-accepted Ethernet standard and represent a three- to six-times improvement for the majority of enterprise sites. The 10-Mbps rate, furthermore, need not be tied necessarily to Ethernet services. This bandwidth, flexibly provisioned without truck rolls, also would overcome both enterprise and carrier market constraints.

G.SHDSL.bis, a new DSL standard, can exploit copper plant's remaining sweet spot. This standard supports a data rate of about 5 Mbps per twisted pair. Companies including Actelis, Hatteras and Aktino combine two pairs (bonding) to get to the 10-Mbps data rate. These vendors' solutions are especially attractive as an upgrade to existing T1 circuits because the necessary clean copper pairs will already be in place. All that's needed is to connect the relative inexpensive terminal equipment onto each end of the existing four-wire circuits. This simple approach, however, is also a weakness in that such point-to-point solutions do not fit well into large ILEC loop architectures that are built around aggressive deployment of DSLAM and DLC equipment. They also lack the elaborate OSS support needed for large-scale deployment--especially in RBOC networks.

Many other copper-based approaches exist to break through the 1.5-Mbps barrier, including DSLAM/DLC systems employing G.SHDSL.bis, IADs (integrated access devices) or TDM muxes, and CPE-based systems such as those from ANDA and Overture.

In addition, many optical Ethernet and PON technologies can approach comparable price ranges, assuming optical fiber can be affordably obtained. Finally, emerging WiMAX/Wi-Fi solutions from vendors such as Proxim, Airspan, Aperto, Tropos and BelAir eliminate cabling of any kind.

Though G.SHDSL technology, as well as several of the alternative approaches, seems well positioned to overcome enterprise and carrier concerns about serving mid-sized locations, strategic and economic factors will determine ultimate success or failure. Key factors will be the priority ILECs place on this market segment compared to the more heavily threatened and regulated consumer segment, the sophistication of vendor approaches to ILEC OSS issues, and the necessary buy-in of major RBOC systems vendors such as Alcatel, Lucent, Nortel and Cisco.

Michael Kennedy is co-founder and managing partner of Network Strategy Partners, LLC (NSP)--management consultants to the networking industry ([email protected]).

COPYRIGHT 2004 Horizon House Publications, Inc.
COPYRIGHT 2004 Gale Group