Can you remove multiple RFI sources at the same time with spatial filtering? Perhaps, but it’s not easy.
Here’s a typical situation on 40 meters above. You can see two signals of interest, S1 and S2, traveling hundreds or thousands of kilometers from different directions. On the other hand, you can see three localized sources of noise or radio frequency interference. N1 is a switched mode power supply from a neighbor to the west. N2 are birdies from a nearby router east of me. Finally, N3 is a mix of broad atmospheric and local noise from many directions. Now, in my current case, neither N1 nor N2 is jamming my desired signals, but this is not always so.
Can I use spatial filtering (phasing) from two antennas to reduce noise without reducing the desired signal strength? Or, can I use phasing to tackle multiple RF sources at the same time? Let’s take a look.
I have three phasing scenarios available. First, is to use (A) a small noise probe against my receiving antenna. Second, I can use (B) two different receiving antennas. And third, I can (C) phase two identical but spaced receiving antennas.
Using a noise probe (A) works best for reducing noise without degrading the signal. You will remember this is because the noise probe doesn’t really pick up much signal, so you end up only phasing noise by 180 degrees. There is not much of an antenna pattern involved.
On the other hand, the (C) phased array approach creates an antenna pattern with at least one null. If the null ends up pointed at both the desired signal and undesired RFI, then you reduce both and may not get an SNR improvement. Your results from (B) phasing two different receiving antennas falls somewhere in between A and C.
Reducing Multiple RFI Sources at Once
In short, eliminating multiple RFI sources at once, e.g. N1 and N2, becomes harder if they are arriving from different directions and you are using an antenna pattern for the reduction. You can use the noise probe approach for better results only as long as the noise probe picks up enough signal from the different sources, which is not always easy as by definition, noise probes are lousy antennas.
Your results are about the same whether you use digital (baseband diversity) or analog filtering. Although, in most cases you have finer control with digital. Keep in mind that directivity is less of an issue above 10 MHz if you are using a wideband magnetic loop.
As for the broadband omnidirectional noise, N3, I find I can improve SNR with any of the approaches, most of the time.