We mustn’t get too excited about the details leaking through from trials of DVB-C2 cable technology, announced to the German press last week by Kabel Deutschland, despite savage improvements in overall throughput. The core design of DVB-C2, using Coded OFDM, like so many modern communications architectures (LTE, WiMAX, 5GHz WiFi, DVB-H, DVB-T), should add 30% of capacity to every cable architecture out there, but its implementation is not so simple. It also has added the LPDC (Low Density Parity Check) FEC which was used in DVB-S2 the satellite standard.
Kabel Deutschland did its field trial just 5 months after DVB-C2 went through ETSI as a European Standard, but tried the technology only on a single channel on an existing network. The results suggest that if an entire 862 MHz cable network shifted to this technology it would deliver 8 Gbps rather than the 5Gbps that is achievable today.
But just how each cable company goes about this is a more complex question.
Kabel Deutschland says that it will go ahead and shift to the new technology starting in mid-2011 and back in March chief technology officers the top seven European cablecos who between them had more than 22 million cable households, declared their support for the DVB-C2 standard.
Right now, to implement DVB-C2 the cable will have to be segmented and two signals will need to be supported, one the existing QAM signal and the other an OFDM signal. Older CPE would not be able to benefit from the improvements, and specifically DOCSIS 3.0 improvements would be short lived and need to be replaced to get the full benefit.
Critics suggest that the OFDM signaling should be implemented at the same time as a change to the Cable Model Termination System which supports DOCSIS and many cable operators may find that it is economically simpler to wait until DOCSIS 4.0 is defined and hope that it also uses the same COFDM architecture.
For networks that want to shift earlier, it would mean running DOCSIS in one separated portion using native QAM, supporting legacy cable TV CPEs with QAM and introducing DVB-C2 (which as we understand it can emulate QAM) side by side, with the duplication of technologies and signals that this would entail. The benefits from bringing it in early are unlikely to deliver dramatic results, about 30% in the channels where it is used, so you might replace 2 channels with 3 OFDM channels. But once an entire system is shifted over to this transmission system, there are much larger gains, potentially in the order of 235% (so from 5Gbps throughput to over 15 Gbps).
There have been concerns since OFDM was first considered as the carrier signal, that simpler strategies like cable splitting, a long time common practice among cablecos, would be more fruitful in the short term ‘ effectively having more, shorter loops, to cope with capacity issues.
In the meantime DVB-C2 receiver chips will have to be built which cope with both QAM and OFDM and these are already under development at several manufacturers. During the field test, a Sony device using a prototype DVB-C2 demodulator from the University of Braunschweig, was used.
Eventually cable systems using OFDM exclusively will have much greater capacity and it is likely that as 3D becomes more commonplace, on perhaps 25% of channels, there will be the need to use DVB-C2 end to end. That might be a convenient stage for Cable operators to begin another round of CPE replacement, for instance with DOCSIS 4.0, which would likely mean in three to four more years.
During the Kabel Deutschland tests DVB-C2 managed to delivery HDTV signals on channel D562, alongside analog and digital services, on a fully loaded Berlin cable network. The Kabel statement talked about the successful transmission of 1024-QAM and 4096-QAM constellations (which are beyond existing QAM equipment) using the new DVB-C2 cable standard ‘ so presumably this is some kind of emulation measure.