Cable Engineers Seek New Standard to Deliver EPON Signals Over HFC
Cable operators are looking to bring greater bandwidth and faster data transmission speeds to cable’s broadband pipes without resorting to costly all-fiber network buildouts. A team of cable technologists is promoting the idea of delivering Ethernet passive optical network (EPON) signals over the industry’s hybrid fiber-coax (HFC) networks in the near future, executives said in interviews and told a recent industry conference. The team of senior engineers from major cable operators and vendors is looking to win approval for their project from the Institute of Electrical and Electronics Engineers (IEEE) this summer.
Sign up for a free preview to unlock the rest of this article
Communications Daily is required reading for senior executives at top telecom corporations, law firms, lobbying organizations, associations and government agencies (including the FCC). Join them today!
The proposed new EPON Protocol over Coax (EPoC) standard would enable cable operators to use EPON, rather than the industry’s current DOCSIS protocol, to send data over their existing hybrid networks without installing more fiber lines. That would allow operators to offer symmetrical broadband speeds as fast as 1 Gbps or possibly 10 Gbps, much faster than they can now provide with DOCSIS 3.0. If the IEEE approves the EPoC project, the new standard could be ready in two to three years.
As DOCSIS 3.0 deployments spread worldwide, EPoC backers argue their proposed standard is needed because the DOCSIS protocol may soon bump up against its bandwidth and speed limits. DOCSIS 3.0 modems can bond eight downstream channels and four upstream channels, supporting speed bursts of up to 320 Mbps downstream and 120 Mbps upstream. These configurations could go as high as 32 channels, at least in the downstream -- enough for speed bursts of more than 1 Gbps. In some recent tests, cable engineers have demonstrated that such high speeds are possible. It’s unclear how much faster speeds the underlying silicon chips in the cable modems can support.
"We need to get ahead of this,” said Jeff Finkelstein, senior director of network architecture at Cox Communications and a strong EPoC supporter. “If we just sit back and wait for things to happen, it puts us at a disadvantage with competition, with chip manufacturers and with the needs of the customers.” Most EPoC advocates believe that cable operators would initially use EPoC technology for commercial customers, who are demanding ever-faster data speeds, higher-capacity services and more of a balance between upstream and downstream capacity. Some think that EPoC could end up getting deployed in residential networks as well. EPoC supporters play up the ability to deploy the technology piecemeal, wherever it’s needed.
Bright House Networks has been deploying EPON technology over fiber lines for business services since 2006. John Dickinson, senior director of network strategy and architecture, said he hopes EPoC will give the company an extra boost as it seeks to deliver more bandwidth in areas not served by fiber. “We are already doing EPON [for business services], so it makes sense to have options there,” he said. “EPoC lets you place fiber where you need it, and leave coax in other areas.” Dickinson also likes EPoC because its goal is to become a worldwide standard that would help drive equipment prices down as manufacturing volumes increase.
Not everybody agrees a next-generation solution like EPoC is needed. Other cable engineers believe that DOCSIS may not run out of gas for a long time, as long as the proper tinkering is done. Chief Technology Officer John Chapman of Cisco’s cable access business unit believes that the cable industry has hardly scratched the multi-gigabit potential of the 15-year-old DOCSIS standard. He foresees the development of cable modems that can bond 24 channels, giving cable the potential for 1 Gbps downstream speed bursts, within the next year. “DOCSIS is just a teenager in many ways,” Chapman said, noting that cable is utilizing just 1-2 percent of its spectrum with the bonding of eight downstream channels. Once operators start using their full slate of spectrum for IP video, he said “we can see how to get to 5 Gbps down and 1 Gbps up."
DOCSIS is far from being a reliable conduit for the new Internet Protocol video services that cable operators plan to offer, Chapman acknowledged. Preparing DOCSIS for managed IP video services -- as opposed to current unmanaged over-the-top video services -- represents “the next technical challenge for our community,” he said. Chapman believes that cable operators will have to set aside about 20 DOCSIS channels to carry a full lineup of IP video services. Some industry experts think that cable providers will probably need to reserve even more channels for the IP video transition.
For EPoC supporters, one chief goal is to make sure that EPoC can work with the Converged Cable Access Platform. CCAP’s a super-dense platform now under development that will combine edge quadrature amplitude modulation (QAM) and cable modem termination system functions in one large box at the cable headend. Ahead of other cable operators, Comcast has issued a request for proposals (CD April 17 p5) for CCAP equipment as it gears up for some small-scale deployments of the technology later this year. EPoC is “definitely complementary” to CCAP, said Vice President Jorge Salinger. “EPoC gives us an alternative to having to deploy fiber to every premise."
Chapman also stressed that EPoC wouldn’t necessarily compete with or replace DOCSIS. Although they would share the same cable spectrum, he said DOCSIS is primarily used for residential data, while EPoC would likely target business-class services, at least initially. EPoC and CCAP should play nice together because they address different market opportunities, even if they might live in different parts of the cable frequency spectrum, Chapman said. “It’s really a usage scenario.”