wbz1030 said:
To clarify - the question is in applications of a local single frequency networks. That is a network of lower powered, medium wave transmitters (as opposed to a single, relatively higher power transmitter) that is optimized for local area coverage.
So to comment on your points brought up thus far,
> (1) Transmission is not instantaneous - at night, skywave would cause the signals to be received at different signal strengths
> and different times. This would create jitter on the digital signal, and if the jitter was bad enough, there is no way for the
> system to stay locked. It is also impossible to time the various skywaves, because every location in the country would require
> the signals to be delayed differently.
This is all true. However, let's consider the network is relatively low powered and the network uses lower efficiency antenna that minimize skywave propagation. The goal is to maximize local area coverage.
> (2) Adjacent channel interference would still be a factor. Unless 1 to 2 frequencies either side of the single frequency were
> vacated, interference would still potentially cause a loss of lock.
This potentially could be a factor at nighttime. Let's assume the SFN is purely operated in an all digital mode which occupies a 10 kHz (or a 20 kHz) bandwidth. Let's also assume the receiving radio has adequate adjacent and alternate channel selectivity. Adjacent channel interference would also be a problem for an analog AM station as well.
> (3) Would the frequency be blank internationally, or just domestically? If not - then unfriendly governments like Cuba could
> still put a station on the air, interfering with digital reception.
Let's again assume a local channel case where external interference (on channel) is not an issue. Cuba has put on stations that had (or continues) to be an issue with analog AM stations as well.
> (4) Interference from natural and man-made sources would still be problematic. One lightning storm in the vicinity, and
> digital transmission is done until the weather passes. Also the same man-made sources of interference - CFL's for example - will
> cause loss of lock.
This is already an issue with AM HD Radio the way it is implemented now. One of the goals of a SFN would be to increase the local signal strength distributed more equally across the intended coverage area than a single (big stick) approach.
> (5) AM is even worse at penetrating buildings than FM. So there would still be no advantage for the office worker.
Again one of the goals of a SFN would be to increase the local signal strength distributed more equally across the intended coverage area than a single (big stick) approach. There may be a better chance, with a SFN approach, at building penetration in the medium wave (AM) band.
First of all WBZ1030, thanks for bringing something more of an intellectual and technical discussion to this board, whether your ideas play out or not. To me it's a welcome change from reading the same anti, or pro-IBOC mantras.
I believe even if one were able to provide a "distributed" or cellular approach to delivering digitally modulated signals in the Medium Wave broadcast band, there would be two hurdles: One that comes to mind is the need to sync and delay all the individual carriers so you don't end up with lost bits due to cancellation from carrier shift and overlap. The other would be just the physics already presented here. Unless one put a sub-transmitter on every floor, of every office building in a market, you would still have problems with reception of Medium Wave signals due to the Faraday effect combined with terrestrial noise at those wavelengths.
Several years ago I got an expermental license, and attempted to build an on-channel booster for an AM station that my friends owned. Obviously we couldn't exceed the allocated RMS for the station, but the idea ibeing that we could fill-in the downtown corridor with additional field strength. Even with the analog signal and the main and booster oscillators locked to GPS, plus audio delay added to the main station to match the booster audio, there was still a +- one cycle shift in the audio at the carriers and audio where the field strength of the main station, equaled the booster. Unfortunately the overlap area was too large to ignore, and in the end became a deal-breaker. Unlike FM or DTV booster stations, at low frequencies and the associated ground wave, just using terrain to create a barrier between the main and booster won't solve the problem. Using the conventional thinking with the existing IBOC algorithms, I suspect decoding the fragile bitstream would be a problem with any frequency shift.
Okay so that is in my view why it wouldn't work, now let's look at how it could work... There isn't any practical way to get around the Faraday cage physics issues of buildings, tunnels, etc., but there may be a way to overcome the overlap shift packet delivery issues. The solution would involve massive Bit Error Correction on the transmit and receive sides, in that the packets would be repeated multiples over time on the transmit side. On the receive side, the demodulating device would have a decent sized cache, (buffer), which wouldn't be able to play the audio out until it all arrives at the device, (QOS). Obviously the latency would not allow this to fit within the traditional live radio model, and would be more of a MP3 and QOS file delivery process.
