The type of radio signal that LightWaveRF and similar devices send is known as OOK (On/Off Keying). This is because they simply switch the signal on and off for periods of time, rather than using amplitude or frequency modulation.
There is an article at Jeelabs on how to do OOK with the RFM12.
I tried this with the LightWaveRF signal from the window switch, which says it operates at 868MHz. I had to add a resistor and wire to the FSK/DATA/nFSS pin on the RFM12B. I did not change the capacitor on the device, so the result I get are not likely to be reliable at short of long distances from the transmitter.
I modified the code written by JGJ Veken to investigate the signal I get from the LightWaveRF window switch. The good news is that is does pick up a signal. However, the signal appears to consist of just one bit! When the window switch is opened I get a pulse of varying length, and then 5 seconds later I get a second pulse whose length is a bit more reliable and averages at about 800 microseconds. The window switch flashes its LED twice when it is opened.
I get a more or less identical signal from the window switch when it is closed. The LED flashes once in this case. Again there is another “bit” 5 seconds later. It fact the second signal is usually 5041 or 5111 milliseconds later.
There is just enough information here for me to detect the window being opened or closed, but I cannot distinguish between the cases. If I had more than one window switch, I would not be able to distinguish between them.
I was expecting to see a packet of about 8 bytes (as with the 434MHz devices), including a unique id for the transmitter. I guess it is there, but the method I am using does not show it.
Things to try next are changing the capacitor on the RFM12B and monitoring the signal with an oscilloscope or equivalent.