In getting signals back to the headend, upstream bandwidth always has been a challenging task for cable operators. DOCSIS 3.1 offers new tools for significantly increasing the upstream capacity.
Just as this light at the end of the tunnel appears, however, the driving need to do so - at least in the short term - is ebbing. It is rather ironic: During the heyday of peer-to-peer (P2P) and file transfer protocol (FTP) traffic, it seemed that the upstream would be a major obstacle for operators. Now, there is little focus on implementing the most aggressive approach to upstream capacity expansion, the migration to the mid-band split.
John Mattson, senior director of marketing for Ciscoâs (NASDAQ:CSCO) Cable Access business unit, suggested that the strain on the upstream caused by peer-to-peer networking and file sharing has moderated simply because laws have made much of the P2P and FTP traffic that dominated a few years ago illegal. The content it carried now is delivered by Netflix and other streaming services.
This is likely a temporary situation, however. While operators don't have to rush advanced upstream capacity into their networks, there is general agreement that it will be necessary in the long run. Once the need does arise, there are significant operational challenges to the mid-band split. Prudent operators, experts say, should think of the mid-band split as they invest in amplifiers and other equipment today.
DOCSIS 3.1 includes advanced modulation and agile channel bonding capabilities that increase upstream capacity without expanding the return path bandwidth. All told, the standard holds the potential of offering 1 Gbps upstream capacity.
Though there is no rush, it is just as clear that the mid-band split at some point will be necessary. "There is going to be a time that demand trends up as people start sharing more video files and that sort of thing," Mattson said. "But it is not there yet."
That doesn't mean that the industry is not moving forward on the DOCSIS 3.1 upstream. In late March, Casa Systemsintroduced a DOCSIS 3.1 upstream module for its C100G and C10G Integrated Converged Cable Access Platforms (CCAP). The company says the module will support upstream transmissions in the 5 MHz to 100 MHz range.
The interlude between when the details of DOCSIS 3.1 are clarified and when upstream demands grow is a good time for operators to plan. Dan Rice, the vice president of network technology for CableLabs, first points out that DOCSIS 3.1 is backward-compatible with previous iterations of the spec. "Operators ought to understand that they don't have to do anything to the upstream to use DOCSIS 3.1," he said. "One of the key underlying principles and objectives was that it should work on existing networks."
Once the move to the mid-band split becomes necessary, operators have several options. They can raise the ceiling - which in North America is now at 42 MHz - to 65 MHz, 85 MHz or even 204 MHz. With the upstream - whatever size it is -Â DOCSIS 3.1 allows the agile bonding of channels of different sizes. This flexibility along with the ability to move a new modulation scheme called orthogonal frequency division multiplexing (OFDM) greatly enhances upstream capacity.
Rice added that a second technique further enhances flexibility and, ultimately, overall capacity: In the past, the upstream transmission strategy relied solely on time-division multiple access (TDMA). Each endpoint got its own period of time to communicate with the CMTS. In DOCSIS 3.1, however, time-division duplexing (TDD) enables a single endpoint to momentarily take over the entire upstream capacity in order to send a short but dense burst of data. Of course, the system must compensate for the data from other endpoints that was superseded when one endpoint took over.
Cable operators must think carefully about expanding into the mid-split. There are a couple of fairly compelling reasons to wait until it is necessary to do so. Raising the ceiling on the upstream may require an enlargement of the downstream to 1 GHz or beyond. Operators must consider the status of filters, amplifiers, legacy set-top boxes, fiber nodes and passive devices in their networks. "That's a very capex and labor intensive exercise and will be done on a needs-only basis, in my opinion," said David Keane-Mirajkar, a distinguished consulting engineer for Casa.
The second issue is clearing out the capacity itself. Like a shopping mall that eyes a particular huge lot on which to build, the current occupants must be relocated or otherwise taken care of. The capacity that the mid-split aims to use likely is being utilized for something else today.
DOCSIS 3.1 downstream capabilities will have a far more immediate impact. In the future, however, DOCSIS 3.1's impressive upstream may prove to be just as important to the industry. "I am not saying that upstream bandwidth demand is not growing," said Jeff Walker, the senior director of CMTS product marketing for ARRIS (NASDAQ:ARRS). "It is. But is it going in the next year or two to grow to need the type of spectrum that DOCSIS 3.1 capable of? No. But it will get there."
Carl Weinschenk is the Senior Editor of Broadband Technology Report. Contact him at [email protected].