Frank, unfortunately I haven't been following this train of thought closely on the thread, so I'll just throw out some random thoughts on the matter of digital baseband in cable plant.
A pure digital baseband cable network has been considered nirvana by the industry for quite some time. The problem being one of history, of which I'm sure you are quite aware, and the evolution has been fairly slow to embrace digital formats of any kind.
Since all content format from the inception of cable has been in analog format, when fiber was introduced, and hence the HFC concept, there were 2 driving forces in place to determine what architecture would be in place.
The first was that everything was in analog format since it was all video entertainment. They looked at going to digital lasers as the phone companies had (where everything is digitized, but in 4 kHz chunks, not 6 MHz channels) and the cost to digitize that bandwidth at the time was prohibitive. So along came a company that demonstrated analog lasers and now the MSO could just reuse its existing modulators to go downstream. So today everything is analog fiber and all digital services are modulated to QPSK or QAM or OFDM (even FSK in some cases dare I say).
But that does not get away from the basic ANALOG nature of the cable plant whether it be 550 MHz, 750 MHz, or 1 GHz. It's all still analog bandwidth, with digital modulation placed on that analog carrier. It's changing the carrier where the answer lies.
The second major factor in place when HFC became viable was that of economics. Payback periods, cash flow (swimming pools, movie stars). Cost factors at MSOs reign supreme.
Can the MSO go to baseband digital carriers? The technical answer to that is yes. The economic answer is probably not for a while. Several innovative suppliers are dipping their proverbial toe in the water. Scientific-Atlanta actually has a transmission system that incorporates a pure digital baseband return path (upstream). And it is cost effective. It is difficult to implement due to legacy applications that want to modulate their digital signals into QPSK or QAM (cable telephony or CM data). It's almost a chicken and egg thing. Once the baseband digital is in place, they can use pure digital signals for transport. In the meantime, they can't and use QPSK or QAM. Until they stop using QPSK or QAM, the operators won't go to baseband digital! Downstream is tremendously more legacy driven, where whole headends would have to be turned upside down to convert to baseband digital.
When provided with the inevitable bandwidth exhaustion question, the standard response is that digitizing the channels, while maintaining the analog carrier, will yield sufficient bandwidth gains to the foreseeable future (at least 3 high quality digital video channels replaces 1 analog 6 MHz channel). And then there is always node splitting, which actually doubles the available bandwidth in a neighborhood each time the node is split into smaller and smaller "pies".
But, as has been suggested in these spaces, true baseband digital service on cable is an attractive possibility that is under development by most major cable suppliers now due to the overwhelming trend in the industry to digitize everything and place integrated services on IP based networks. We now have the paradox that the additional expense of doing the additional digital modulation of QAM or QPSK to place these services on an analog carrier is getting restrictive.
We'll soon have all digital voice, video and data running over the legacy analog carriers due to architectures planned and designed when there was nothing more than one-way video entertainment to consider. And the architecture of cable plant today is still designed and engineered to optimize that analog video service. The advances that we've seen in recent years that incorporate SONET transport rings and metro-DWDM narrowcast systems are a reaction to the proliferation of advanced digitally based services that demand more than the original analog laser carrier systems could handle.
So is the dream of a pure digital baseband cable network possible? Absolutely. Just keep in mind that industry drivers when you contemplate when such a system could be implemented.
PS - as you can see, I've got a few spare minutes these days, don't know how long it will last. I've been enjoying some breathing space. Exciting and hectic these days what with cable telephony booming and the transition to IP telephony on cable networks just now coming to fruition. Some I'm caught between 2 worlds, having to serve both. The talent pool in the industry that knows both circuit switched cable telephony and IP based cable telephony is virtually nonexistent, so guess who they come to?
DT |
