# How do you null interference with beamforming?

I am trying to understand a text about beamforing :

Classic beamforming toy example. Single spatial stream, 2x2. With one interferer, receiver alone can null it. With two interferers, need to coordinate nulls in both receivers and transmitters. For example, the nulls for the transmitters to receivers are (1,2), (2,3), (3,1) and for the receivers to transmitters are (2,1), (3,2), and (1,3), thus using all six degrees of freedom to cancel the six interference paths. With three interferers (four user pairs), we lack sufficient degrees of freedom to cancel interference. Thus, we resort to orthogonalization of channels.

What do we mean by nulling in this case? Are interference coming from the antennas inside the 2x2 system? is the case with two interferes a 3x3 system and the next one a 4x4 system? What is a user pair, is it an antenna in the MIMO setup? What is a degree of freedom in this case and why do we have only six of them?

• Welcome to EE.SE. Could you add a little into your post explaining what "beamforming" is (for the rest of us)? You should also reference the text you mentioned so that others can get context if required. Aug 23, 2016 at 21:05
• Welcome to EE.SE. Your question is a wall of text. It's very hard to digest. Try to reduce the number of questions to 1. Try to very clearly structure your text visually – visually insert a boundary between definition and example. Aug 23, 2016 at 21:08
• @NickAlexeev I don't think MIMO is accurate here. By my understanding/guess of the question, I think the OP is using a 2-dimensional array of 4 antennae sending/receiving a phase-corrected copy of the same identical signal in order to "steer the beam" of an RF transmission using the phase-induced constructive/destructive interference. -- Might be a better-place question over on the Amateur Radio stack tho. Aug 23, 2016 at 21:10
• @RobhercKV5ROB don't thin that the amateur radio SE would have significantly more knowledge than this SE site filled with Electrical Engineers :) The problem really is that karb22 should clearly define things, then answers might be derived really easily. The way karb22 underdefines the problem would require anyone to write a 20 page report on how to model such systems. Aug 23, 2016 at 21:21
• Thanks for the quick comments. This is an excerpt from a proposal draft (printed) that my professor has given me and left me with for a month to figure it out. I just started graduate school and I'm trying learn these things on my own for now and before the semester starts. Aug 23, 2016 at 21:21