By Stephen Hardy
The SCTE Cable-Tec Expo is annually the best place to get a reading on what technologies cable operators in the U.S. plan to deploy in their networks. That was again true October 21–24, when the cable industry gathered for this year's edition in Atlanta. While much of the technology discussion revolved around the imminence of such HFC-specific advancements as DOCSIS 3.1 and the Converged Cable Access Platform (CCAP), passive optical networks (PONs) also received a fair amount of attention. However, it isn't quite clear whether that attention was focused on the right place.
The assumption among most observers with whom I've spoken is that cable operators will deploy EPON if and when they finally get around to mass adoption of optical access technologies. Certainly CableLabs is operating along this line. The U.S. cable industry's primary standards group has finished the first version of its DOCSIS Provisioning of EPON (DPoE) specifications and has its eye on the IEEE's EPON over coax (EPoC) standards work. A CableLabs spokesman described during one session how cable multiple systems operators (MSOs) might employ EPoC alongside EPON and point-to-point Ethernet networks, particularly to supply business services.
All that is fine, except for one detail: If a cable operator has deployed a PON already, it's more likely that the infrastructure is based on GPON, not EPON.
Why is this the case? For one thing, at 2.5 Gbps, GPON offers greater downstream speeds than EPON's 1 Gbps. It's also the preferred PON of the telcos against which these cable operators compete, so it's more likely to create a level broadband access playing field.
That would appear to imply there's a disconnect between CableLabs and its constituency in terms of which PON technology to pursue for standards work. In fact, representatives of some vendors on the SCTE Cable-Tec show floor who support DPoE have said they've received inquiries from cable MSOs about the feasibility of "DOCSIS Provisioning of GPON."
But that apparent disconnect may not be as significant as one might fear. As mentioned previously, most of the EPON-related efforts are targeted at business applications, while many if not most of the GPON deployments address residential service requirements. And with the advent of DOCSIS 3.1 and its 10-Gbps capacity promise, cable operators won't have to turn to GPON fiber to the home to keep up with Verizon, Google Fiber, and other competitors who leverage optical access technologies. When operators do need fiber access, DPoE offers an option that's more economical for dense applications than RF over glass (RFoG) while maintaining compatibility with DOCSIS, so operators won't have to use separate operating systems for their HFC and PON infrastructure.
So there will be less reason to install GPON rather than EPON. The question then becomes whether the cable operator in question wants a PON-based approach at all. If that answer is yes, EPON will likely become increasingly popular in cable-operator networks. And CableLabs will avoid having to wipe egg off its face.
Archived Lightwave Issues