This new platform separates functions that have always been performed at the headend, which is going to allow for a more robust data network. Today, the cable headend is the place where all video is inserted, all cable management is done, where the QAM modulation and RF Modulation is performed, and most importantly where the CMTS (cable modem termination system) function is done.
The distributed CCAP allows these functions to be separated and geographically distributed as needed throughout the cable network. The main benefit of this is that a cable operator will be able to push pure IP to the fiber nodes. Today, the data path between the headend and the neighborhood nodes needs to carry two separate paths – both a video feed and a DOCSIS data feed. By moving the CMTS and the QAM modulators to the fiber node the data path to the node becomes a single all-IP path that contains both IPTV and IP data. The new CCAP node can then convert everything to RF frequencies as needed at the node.
We’ve been expecting this change since for the last few years Chinese cable networks have implemented the distributed network functions. Probably one of the biggest long-term potentials for this change is that it sets the stage for a cable company to offer IPTV over DOCSIS frequencies, although there is more development work to be done in this area.
There are several immediate benefits to a cable system. First, this improves video strength since the TV signals are now originating at the neighborhood nodes rather than back at the headend. This will be most noted by customers who are currently at the outer fringes of a cable node. The change also will boost the overall amount of data delivered to a neighborhood node between 20–40%. It’s not likely this mean faster speeds, but instead will provide more bandwidth for busy times and make it less likely that customers lose speed during peak hours. Finally, it means that a cable company can get more life out of existing cable nodes and will be able to wait longer before having to ‘split’ nodes to provide faster data to customers.
Cable companies are not likely to rush to implement this everywhere. It would mean an upgrade at each node and most cable companies have a node for every 200–400 customers—that’s a lot of nodes. But one would think this will quickly become the standard for new nodes and that cable companies will implement it over time into the existing network.
This is the first step of what is being called the IP transition for cable companies. Most of my readers are probably aware that the telcos are working feverishly towards making a transition to all-IP. But cable companies are going to want to do that for a different reason. There is a huge amount of bandwidth capability on coaxial cable and if the entire cable network becomes IP from end-to-end then the huge data capacity in the cable network would be realized. Today cable companies use a broadcast system where they send all cable channels to every home and they then provide data services on whatever bandwidth is left. But in an all-IP system they would only send a customer the channels they are watching, meaning that most of the bandwidth on the system would be available for high-speed Internet services.
So think of this as the first step in a transition to an all-IP cable network. There are a number of additional steps needed to get there, but this pushes IP out to the neighborhood nodes and starts the transition.