We're getting close to one of the North American cable industry's major events -- the Western Cable Show -- to be held across the street from Disneyland in Anaheim just after Thanksgiving. Everyone in the industry tends to synchronize announcements with the Western show (and with the National show in late spring), so we've seen a flurry of cable announcements in the past month.
Tracking an industry over a long period helps us understand how seemingly isolated events form coherent trends. We've put together some recent announcements to give our readers a clearer picture of their context and the trends we see from them.
"Cable ISPs" have served their purpose
In the early days of cable modem service, just about everything had to be invented. Cable modem standards (DOCSIS) didn't exist; cable systems had a lot of technical problems; there was no pricing structure or marketing materials; specialized broadband backbones and a delivery architecture needed to be developed -- and the cable guy had no idea how to install something on a user's PC. The cable industry tends to operate like an extended family, and the notion that they'd each have to invent everything to roll out data services didn't seem sensible -- so the concept of "cable ISPs," like @Home and Road Runner, was born. Reporting on the formation of @Home in November 1995, USA Today said that "The hottest new cable service has nothing to do with TV".
A lot has changed since 1995. We have DOCSIS cable modem standards, with over 40 companies now making compliant products. Many of the early problems have been addressed and solved, and specialized companies are dealing with the outstanding ones.
We think we've now transitioned thru the stage in which "cable ISPs" were needed to develop, aggregate and dispense the specialized knowledge and services to run high speed data over cable; it seems appropriate that the functions these companies once performed for multiple MSOs are now reverting back to the individual cable operators.
MSOs making progress toward support for multiple ISPs
While cable operators continue to fight the idea of "open" access, they're making progress in enabling multiple ISP access over their high speed lines. Four of the largest (AT&T Broadband, Time Warner Cable, Comcast Cable Communications and Cox Communications, Inc.) are doing trials or rollouts. However, the pace varies considerably.
Time Warner has made the most progress of any MSO in rolling out multi-ISP access. Under the terms of the AOL/Time Warner merger agreement, TWC is required to bring on at least one nonaffilated cable broadband ISP service before AOL begins its high speed service; they are required to add two more nonaffiliated ISPs within 90 days. EarthLink has already launched on 16 AOLTW systems, and has said it plans to steer its broadband customers toward cable rather than DSL, since cable provides larger profit margins. AOL has also launched their broadband service on Time Warner systems in Raleigh and Tampa.
Despite cable operator fears about multiple ISP access, it's interesting that (according to Cable Datacom News), "Time Warner Cable posted its strongest quarter ever for cable modem subscriber additions, adding 252,000 customers in Q3, an increase of 11.5 percent over its Q2 total." ( www.aoltimewarner.com )
Cox has launched a six-month trial of EarthLink and AOL high-speed services over its El Dorado, Arkansas system -- where Cox already has its own ISP, called Cox Express. ( www.cox.com )
With all the turmoil at AT&T Broadband, including management changes and whether it will be sold or spun off, there has been little news regarding its progress on its multi-ISP program, Broadband Choice. ( www.attbroadband.com )
DOCSIS 1.1 will enable new services and revenue
If the events of September 11th hadn't cast such a pall over everything, we would have opened a bottle of champagne in our broadband home on September 27, 2001. That was the red letter day that CableLabs awarded certification to the first DOCSIS 1.1 cable modems. Why do we view this as such a big deal? Five years ago almost to the day, we gave an invited talk at CableLabs promoting cable's revenue opportunities from small business and communications-based services, which required support for quality of service (QoS) in cable modems. Unfortunately, the first cable modem standard (DOCSIS 1.0) provided only "best-efforts" service, so these services had to wait. ( www.cablelabs.com ) ( www.cablemodem.com )
The DOCSIS 1.1 spec adds key QoS capabilities, including bandwidth and latency guarantees. It permits operators to offer many different services over the same cable bandwidth, favoring some services such as primary telephony over others such as "best efforts" Web browsing. Now that CableLabs has certified 1.1 modems (from TI and Toshiba) and granted 1.1 "qualification" status to cable modem termination systems (CMTSs) from Arris and Cadant, the door is finally open for the new services and revenue streams we have been advocating for years. It enables MSOs to deliver services such as IP voice, streaming media and tiered data services over their broadband access networks, and to offer business customers "committed information rate" services similar to those telcos deliver with frame relay. ( www.ti.com ) ( www.toshiba.com ) ( www.arrisi.com ) ( www.cadent.com )
DOCSIS 1.1 is a key element in the PacketCable specifications for IP telephony over cable. Cable companies view QoS as a requirement for primary telephony, and PacketCable is built on DOCSIS 1.1.
DOCSIS 1.1 should become the norm from this point forward. As 1.1 modems and CMTSs from the leading vendors achieve certification, we expect that MSOs will deploy them widely. Until MSOs add 1.1 CMTSs, they can use 1.1 modems with existing DOCSIS 1.0 equipment. 1.1 modems will also be compatible with future DOCSIS modems, such as the recently announced DOCSIS 2.0.
Next generation head-end equipment supports new services and higher penetration
The first generations of cable modem head-end equipment were devised to enter the market quickly and handle low penetrations of best-effort data services. As these products gained market share, start-ups - such as Broadband Access Systems, RiverDelta Networks and Pacific Broadband Communications - took on the challenge of supporting multiple services and service providers while increasing port density, reliability, and usable bandwidth. The recent news that Juniper Networks will acquire Pacific Broadband Communications for $200 million in stock continues the consolidation of start-ups into established players. ADC Telecommunications acquired Broadband Access Systems last year, and Motorola has just completed its acquisition of RiverDelta Networks. ( www.basystems.com ) ( www.pbc.com ) ( www.juniper.net ) ( www.adc.com ) ( www.motorola.com/broadband )
Meanwhile, the cycle continues as new companies once again address the next opportunities and challenges. These include Cedar Point Communications, whose products are intended to address MSOs rolling out telephony services, and others like Narad Networks and Rainmaker Technologies who are working to dramatically expand available bandwidth for all the coming services. ( www.cedarpointcom.com ) ( www.naradnetworks.com ) ( www.rainmakertechnologies.com/ )
MSOs getting tools to automate service management
With Excite@home going under and the need to keep ahead in their race with DSL, cable operators are adopting additional tools to simplify and expedite provisioning. Although DOCSIS 1.1 modems provide MSOs the opportunity to offer tiered services, it also brings the need to manage additional complexity.
In the evolving environment, operators will need to manage IP addresses, maintain several levels of service, allow for dynamic changes in service profiles and throttling of bandwidth level, support real-time activation of new subscribers and ISP selection for multi-ISP access. Emperative, one company targeting this part of service management, has released a new IP address management system targeted to these problems. ( www.emperative.com )
Plant reliability may be a stumbling block
We've been concerned for many years about running critical services over cable networks, which typically have many places where a single failure can impact tens or hundreds of subscriber homes, and are usually not monitored to alarm for failures. It's great that DOCSIS 1.1 and PacketCable enable telephony services, but it's not great if the operator can't isolate plant problems, measure congestion, or target capacity expansion. By the end of this year, one start-up company, Stargus, is promising to launch its DOCSIS-based network software management solution to address those needs. Stargus will take advantage of the tons of data produced by DOCSIS devices but presently largely unused. The Stargus staff includes several leaders of the DOCSIS effort, so they're very knowledgeable to undertake this. Cable operators haven't previously seen much need for this category of product, so the big question will be whether they're prepared to spend money for it now. ( www.stargus.com )
For more information on the Western Cable Show, see http://www.calcable.org/westernshow . We'll be there, and will provide additional perspectives in the next issue of BBHR.