DENVER -- Cable Next-Gen Technologies and Strategies -- If there were any questions about how quickly DOCSIS 3.1's development has progressed, Comcast Vice President of Access Architecture Jorge Salinger has laid them to rest.
Speaking here on Tuesday, Salinger announced that Comcast Corp. (Nasdaq: CMCSA, CMCSK) is now testing DOCSIS 3.1 in the field. "The target for us is to be in the field establishing network readiness in 2015," he said. "Our overall goal is to be able to deploy DOCSIS 3.1 and gigabit-per-second in a broad scale starting in 2016."
The field tests mean that the largest US MSO has orthogonal frequency division multiplexing (OFDM) signals running over its existing HFC plant. The cable industry's shift to DOCSIS 3.1 marks the first time that the DOCSIS specification is breaking free of its old 6MHz and 8MHz channels and moving instead to OFDM subcarriers. "We have OFDM signals running in one of our networks, in one of our headends," Salinger said.
Because there is no certified DOCSIS 3.1 hardware yet, Salinger added that Comcast is relying on specially designed equipment for its current tests. The cable operator has D3.1 deployed to select employee homes, and Salinger reported that the initial results are promising. Comcast would like to get DOCSIS 3.1 commercially deployed as quickly as possible in order to roll out new gigabit broadband services.
One of the bigger selling points for DOCSIS 3.1 is the fact that cable providers don't have to upgrade their cable plant to enjoy the benefits of greater network efficiency. However, when Cisco Systems Inc. (Nasdaq: CSCO) CTO of Cable Access and Cisco Fellow John Chapman joined Salinger in a panel discussion on D3.1, he made the point that cable operators should consider upgrading their HFC networks anyway to gain the full benefits of the technology. For example, he suggested it would make sense to start planning now to extend the upstream path.
"I would actually encourage early upgrades," declared Chapman, who is considered one of the fathers of the DOCSIS specs. "Every time you go out to segment the plant right now, why not start planning for 85MHz return? Even if you're not deploying anything at 85MHz for the next year, it takes a while to upgrade the plant."
Chapman noted that many European cable operators, which contend with less out-of-band traffic than their US counterparts, are already looking at expanding to 200MHz for the return path. There's also serious consideration in Europe about upgrading to 1.2GHz in the downstream.
In the US, most cable plants operate at 750 or 860MHz, with a few extending up to 1GHz. Chapman, however, believes that ultimately it will be possible to upgrade beyond 1.2GHz capacity to 1.5GHz, and even 1.7GHz.
As for deployment of DOCSIS 3.1, the immediate-term challenges include the development of new installation and maintenance tools, as well as training for cable engineers and technical operations staff. Salinger noted that unlike with DOCSIS 3.0, D3.1 means that cable modems will no longer operate at a fixed modulation. Some cable modems in a service group could be using 1024 QAM, while others could be running on 4096 QAM. New tools must be able to identify which modulation profiles work best where and understand how to manage the modulation diversity.
On the training issue, Salinger said that Comcast has started training between 500 and 1,000 of its people for DOCSIS 3.1, but that the company needs to push those numbers to between 20,000 and 40,000 people. Comcast plans to start commercial deployments of D3.1 in 2016, and there's a lot of training that needs to take place between now and then.
Meanwhile, CableLabs Director of Network Technologies Belal Hamzeh noted on the same panel that the DOCSIS 3.1 certification program is progressing well. CableLabs has already conducted three equipment interops, with several more planned for the near future. The organization's goal is to be ready to start officially certifying equipment by the middle of May. (See First DOCSIS 3.1 Interop Test Complete.)
Across the board, there continues to be a lot of optimism about the exponential benefits that DOCSIS 3.1 can provide, including its ability to extend the life of the DOCSIS platform as a whole. As several panel members joked, maybe DOCSIS 3.1 should have been granted a different name: DOCSIS 4-Ever.
— Mari Silbey, special to Light Reading