There are several potential sources of noise (unwanted signal). The first thing you must do is to acquaint yourself with the stages in your system at which this noise can enter.

First, I'd recommend you start with the easy ones. Have a look at your power supply. If your power supply isn't "clean" then you'll have that same noise everywhere, particularly in the RF gain stages, where it's likely it will be amplified. The MCU circuitry can tolerate a fair amount of noise, but the RF stages can't. Be sure that the RF transmitter and receiver have significantly less than 5 mV of noise at their power inputs.

Second, it can migrate from your MCU circuitry, via RFI, to your antenna. Shielding the antenna subsystem from radiated noise from the PSU and MCU circuits can save you considerable headache. Check the antenna wiring, too, and ensure there's a good solid ground where it is required.

Third, look in the environment near your system. If there is equipment, even test equipment that has fans, etc, they are a potential source of noise. Distance is probably the easiest way to mitigate that, but placing grounded metal screens (Google for information on "Faraday cage") around the test equipment, particularly PC's, can help considerably and doesn't require much in resources. Switching power supplies sometimes radiate quite a bit and do it in the range of popular baud rates.

One thing that will help greatly, if the equipment is available is to attach a spectrum analyzer to the point at which you observe the noise that you suspect is causing your problems. If it is, in fact, there, you should be able to identify its source from its spectral composition. I doubt you'll have to go that far.

One thing that may help avoid "barking up the wrong tree", so to speak, is to ensure that the communication works without the RF link before trying it with the RF link.

RE