The CTIA and CES have produced a white paper outlining how to extract 180MHz of TV-broadcast spectrum without impacting TV quality or coverage, and it's worth reading.
The Cellular Telephone Industries Association (CTIA) and Consumer Electronics Association (CES) co-authored the paper (pdf) which proposes replacing centralised TV transmitters with a network of small, low power stations which would work with existing set-top boxes and free up between 100 and 180MHz of spectrum - spectrum which the CTIA's members would love to get their hands on.
Proposals from the CTIA often verge on the hysterical - such as predicting the downfall of American industry if more radio spectrum isn't allocated post haste. Even this one can't help pointing out that only ten per cent of Americans rely (exclusively) on broadcast TV. But after the usual rant, the document sets out a surprisingly reasonable plan for shrinking TV spectrum usage, though a few technical barriers do remain.
Broadcast TV is based around huge transmitters, normally atop massive towers, which pump out enormous signals that can be picked up 40 miles away. Frequencies can be reused, but only for transmitters at least 80 miles apart - actually much further due to variability in signal propagation. So in the USA there are around 1800 towers, which blanket the country with between 15 and 20 6MHz-wide channels (to carry one analogue channel or a digital multiplex) despite the fact that there are 49 slots available.
That disparity is down to the "near-far problem", whereby the strength of signal necessary to reach 40 miles or more, causes the channel to leak into neighbouring channels for users sitting beside the transmitter. That reduces the channels which can be used, and the CTIA reckons it's using a sledgehammer to crack a nut.
The new plan has lots of little transmitters, working on the same frequency to provide the same coverage and with synchronised transmissions. The significant problem here is "ghosting", whereby the TV receives the same signal twice with a tiny time delay between the two based on differing distances to the transmitters. Analogue TV has always had this problem, which literally results in a ghost image appearing slightly offset from the desired picture. But this is generally down to reflected signals bouncing off walls and suchlike, rather than picking up two signals at the same time.
The CTIA argues that modern set-top boxes are so armoured against receiving bounced ghost signals that they can cope with one coming from a different transmitter. Modern boxes accept the strongest signal and are capable of disregarding the others.
That assertion is the weak point, but if the CTIA has its sums right then there's little reason why the system wouldn't work. The group reckons it will cost less than $2bn to built the new transmitter network (based on 15-20 new transmitters for each one currently operating), while it estimates the value of the frequencies released at between $36bn and $65bn. We'd lean towards the lower estimate, but even if it's half that the plan makes economic sense.
Not that access to more spectrum is the only advantage CTIA members would see from the plan. All those new transmitters will need homes, and could easily be bolted onto cellular towers - reducing costs if not providing revenue for the operator. Let's also not forget that that plan would substantially reduce the available White Space which can be used without paying network operators a penny - so it's good news all round.