Smarter transmitters, lower energy: dynamic FM radio networks

Most people never think about the transmitters that bring radio to their homes, cars, and devices. Yet keeping those transmitters running requires a significant amount of electricity. Across a national broadcast network, even small reductions in power consumption could lead to substantial energy and cost savings.

Our research has explored whether weather forecasts could help with this. By predicting how atmospheric conditions affect radio signals, we investigated whether some transmitters could safely operate at lower power levels without affecting coverage.

There are over 200 transmitters broadcasting radio signals - you might have seen the aerial masts on top of hills. All these transmitters use a lot of electricity: over 25 gigawatt-hours per year - a good ten thousand times what you’re likely to use in your home. The 13 biggest transmitters use about a third of that electricity.

Despite all this power, there are areas and stretches of road where signals don’t reach. Smaller transmitters fill in those gaps by relaying the high-power transmissions.

Are savings possible with FM transmissions?

The cost of electricity is high, and it makes sense to use as little of it as possible while reducing the impact on the environment. But high power is needed for transmission because reception conditions are not always ideal. For instance, anticyclones bring fine weather, but also bring interference from distant transmitters. That means enough power should be used to overcome this interference and allow for good reception. However, under normal conditions, could we turn down this built-in contingency power? With modern technology, we could even do that remotely, without the listener being aware!

The idea is shown in the pictures below. In the first picture, the green area shows where good reception of the big transmitter is possible, and the blue areas the same for the smaller transmitters. Because of the contingency built into the system, there is some overlap of the green and blue areas. Overlap indicates good reception of both the big and smaller transmitters. If we halve the power of the big transmitter, the green area has now shrunk to the yellow area. However, there is no problem with that, as the blue areas still overlap the green, reception is maintained, and we save electricity!

In the second picture, interference is present, and the yellow area has shrunk further; there are areas no longer served. The answer then is to return the big transmitter to full power. Coverage is now restored, and is the same as it would be without our energy-saving attempts. Note that the smaller transmitters always remain at full power.

You might wonder about the bother of retuning your radio as you move between transmitters - something that could need doing more often as a result of our power control strategy. Fortunately, car radios deal with this on your behalf. The same system that provides the in-car traffic warnings also decides which is the best transmitter to tune to, and will make the radio select it automatically - without you knowing it. And the radio has yet further tricks up its sleeve to ensure you have the best possible listening experience when conditions are less than ideal.

Testing and results

To test the feasibility of this ‘dynamic’ power control approach, we needed to carry out some research. We built half-a-dozen crates of equipment, aptly named ‘BoB’ - ‘bag of bits’. These could be installed at low-power (relay) sites around the country. Despite appearances, the BoBs are quite clever and can measure interference (among other things) even when this is at a low level.

The BoBs generated a large amount of data: a set of measurements every five seconds over periods of several months. We recovered this data and compared it with weather reports. Using a weather forecast to set the power of the 13 biggest transmitters once each day, we found they could safely operate at half power for 70% of the time. Doing so would save between two and three gigawatt-hours per year. Although this forecast-based approach is cautious, it is still possible for mistakes to be made. If interference occurred at a low-power site despite the forecast, the associated main transmitter could return to full power.

A practical system would require suitable monitoring receivers at the low-power sites. However, those sites already use receivers to rebroadcast signals from the parent high-power transmitters. Modern receivers are ‘software-defined’ and can make the relevant measurements. With increased confidence in the approach, transmitter power could be adjusted almost instantly, depending on the measured interference, giving even greater economy.

Even if the ‘dynamic’ power control approach were not to be implemented, the monitoring receivers would still perform a valuable function by warning the broadcaster of problems with their network. We have mentioned the capability of measuring interference levels, but the same technique can warn of other misdemeanours which could upset the listeners.

While this research focused on FM radio, it forms part of a broader effort to improve the efficiency of broadcast transmission. We’ve already investigated approaches for DAB (here, here and here) and AM services, and hope to make similar progress with digital terrestrial television in future, maintaining the reliable services audiences expect.