# Forum

## Spatial Multiplexing

1 posts by 1 authors in: Forums > CWNA - Enterprise Wi-Fi Admin
Last Post: April 15, 2010:
• Unfortunately in RF, there are several concepts which are extremely difficult to explain in terms that we can physically relate to. Some we can, some we can’t.

For example, most of us can understand the concepts of EIRP [ isotropic radiator, the Sun analogy etc ].

However, when we start getting into things like Fresnel Zones etc, it starts becoming very tricky. If anybody  remembers doing line-ups on point to point FM microwave systems, how do you explain a Bessel Zero  First Carrier Null ? Not easy to do,at all.

http://en.wikipedia.org/wiki/Spatial_multiplexing

Not pleasant. The warning should not  only apply to that article in particular but to the whole of MIMO theory !!

If you are of a mathematical bent and remember matrix algebra and summation theory, “Next Generation Wireless Lans” by Perahia and Stacey gives some great details.

Many of the physical analogies break down or become too simplistic.

You have probably read that MIMO  [ Multiple In Multiple Out ] systems use a number of antennas which have a particular spacing pattern in terms of the wavelength of operation. Multipath is useful so that the signals which arrive at the receiver are said to be de-correlated. In other words, for most of the time, there is no set phase relationship between one wave and another. Sophisticated electronics help make sense of the received signals.

The key to a lot  of this is the intelligence of the digital signal processors which .11n systems can use. These are extremely sophisticated devices capable of looking for patterns in multiple signals.

Imagine that we have a little box with a tiny guy in it [ an analogy …..now I’m doing it !! ]. This little guy can speak German, French and Italian to a very high level. He also has a brain like a computer. Now we turn on a German radio station, a French radio station and an Italian radio station at the same time. To us it sounds like a jumble, but to him, he can identify individual patterns and translate all three streams. This is something like the way a digital signal processor works.

It’s frustrating wanting to know how some of this stuff works, but unfortunately a detailed mathematical analysis is required to get a full understanding.

There is a good basic explanation in the CWNA exam guide.

A similar situation existed [ nothing new under the sun with 802.11 ] when we had point to point microwave using say half a dozen channels operating spread spectrum at the same frequency and at the same time and on the same polarization.  Math is really [ unfortunately ] the only way to TRULY understand it.

Beware simplistic analogies in this area [ including little men in boxes  !!].

I read through the mathematical analysis papers from time to time and get what I think is a fair undestanding of what is going on from a DSP point of view.

A week later…poof !! … it’s all gone.

There are some good animations floating around for some of this stuff. Texas Instruments used to have really good material on DSP systems. I had one of their books years ago, but can’t remember the name.

Now I just call the .11n receivers  “magic boxes ” and be done with it.

Dave

Page 1 of 1
• 1