Vyyo 3 GHz: Bigger Pipes for Cable

Even back in 2005, before the YouTube, user generated video wave hit, cable operators were thinking the amount of bandwith they would be delivering in linear fashion would shrink, while on demand services would grow. Over the top Internet video only magnifies the trend. So it is no surprise that cable operators are looking at any number of ways to boost bandwidth without ripping up their hybrid fiber coax plants and going fiber to the home. After all, the consistent cable refrain is that FTTH is way too expense and HFC is much better.

Cablers are looking at signal compression and decoding, upgrades to 1 GHz bandwidth, creating smaller fiber-served neighborhoods, converting to all digital signal formats or overlaying new electronics on the existing plant to make use of non-traditional frquencies. Vyyo is one of the suppliers of an overlay system, and says it can boost downstream bandwidth by about double, and upstream bandwidth by 10 times, for something on the order of $125 investment for every home passed by the cable network.

That's the good news. The bad news is that every time an engineer adds an active element, serially connected to another active element in a network, reliability necessarily takes a hit. So the good news is that the Vyyo overlay dramatically increases bandwidth, at the cost of reduced network reliability.

If the 3 GHz Spectrum Overlay adds six elements in series on each fiber leg, then the reduced reliability can be calculated by multiplying the reliability of each single device (a number less than one)by the reliabilities of each of the other devices, to get the impact on system reliability. Obviously, multiplying numbers less than one by other numbers less than one can only go in one direction: down.

If one assumes a reliability of 100 percent--a device or system that never fails--you understand why telecom engineers are so fixated on 99.999 percent reliability for the entire network. Never mind that the goal is virtually impossible to meet for some services, as the terminal devices have reliabilities low enough to drop end-to-end performance below the targeted "five nines."

But that's why there is such emphasis on the telecom world on passive optical access networks. Telco outside plant technicians and engineers know just how labor intensive a system can be when it has many active elements in it. For the cablers, perhaps the overlay tradeoff is worth making. But it will come at a price beyond the capital investment, in the form of reduced network reliability. That's just the physics and math.