I went through the same process the OP did when I had to learn SM. Most of the papers on SM out there start with basics and then move to the math. Well, that wasn’t going to work for me. I had to speak to hundreds of people on the topic and there weren’t going to be many PHDs in the bunch.
So, let’s break it down.
- SM is designed to send multiple streams of data over the same frequency simultaneously. The only way to do this is to have uncorrelated signals. An example of an uncorrelated signal is multipath. Multipath is where one signal is reflected in such a manner that its reflections arrive at the receiver out of sync with each other. With 11n and SM, this is (almost) required. Say it with me: “Multipath is good”.
- Here is an example to back this up. If you took two “regular” (Ok, I’ll say it “Non-Ruckus”) 11n devices and had them connect to each other in free space, they will NOT be able to spatial multiplex with each other. So why is it that a Ruckus AP could SM and “others” cannot? Read on.
- One example we’ve used to achieve uncorrelated signals is with multipath. In free space, (say, outdoors) multipath really doesn’t exist. So, can we SM outside? Yes, but we have to find another way to de-correlate the signal. This can be done with polarity.
- Polarity is the orientation of a wave. Vertical polarity is when the waves go up and down like ripples on a pond. This is the polarization of all “stock, rubber ducky” antennas. The other type of polarity is horizontal. This is when the wave travels side to side like a snake in the desert.
- The key to SM with 802.11n is to be able to de-correlate the signals everywhere the STA will connect. This is one of the faults with many 802.11n APs is that they can’t control the RF to any degree. The main problem is with consistency. 11n is fast, but unless you are in the perfect “multipath” room, your performance can change with location, and even device orientation. Control of the RF medium is more important than ever.