I was doing some reading last night, and discovered something very interested. I was reading on Shannon's Capacity theory (I have read up on it before, but never compared it to real tests) Here's a link to a good video that explains it: https://www.youtube.com/watch?v=ancDN11C2vg
I started doing some math on the performance I got with various tests, to see how my real throughput compared to the possible maximum throughput.
I was looking at my tests up on the mountain from yesterday, and discovered that they were at the upper limit of maximum possible bandwith per Shannon's capacity theory.
For example, Band 2 test yesterday:
Note that the SINR is about 20.
Speed test from band 2.
Now do the math and guess what? For a 20mhz channel with a SINR of 20 the maximum throughput possible is.... 87.8 mbps!
(You can do the calculations here yourself if you don't believe me: http://www.rfwireless-world.com/calcula ... lator.html
This also happened with my band 12 test up on the mountain. The ones outside at my house, I ran them through the calculator and they still had a lot of room to improve. If I'm understanding this correctly then, I think this means a few things.
1. The only way to improve my bandwith (at least in the tests at the top of the mountain) is to get a better signal to noise ratio.
2. The antennas were working at full capacity and efficiency if you think about it. The only way to improve from this test, would be to have completely clear LOS (this was close, but you all saw from the pictures that it wasn't quite).
3. Signal to noise ratio is everything. I've noticed this in other tests as well. There's a few reasons for this. (Improving signal strength is a good idea, but SNR is much much more important)
A. LTE modulation rates increase as SINR ratio improves, so you can get more bits per symbol and more throughput. See this link: https://www.researchgate.net/figure/Mea ... 2_47330409
. It's not the greatest diagram, but you get the idea.
B. The more narrow the bandwith, the higher SNR one will have. Why is this? Because the less spectrum you are viewing, the less noise you will have. This is why when I'm using both bands the signal to noise ratio is much lower than when I'm isolating the modem to one specific band.
C. I've seen this on my ubiquiti nanostations. They are 2.4ghz, and if I run a 40mhz wide channel, they pick up tons of noise, but if they are only 10mhz wise, the SNR improves greatly.
4. I ordered some B2 yagis yesterday, so it will be interesting to see if they can improve on the SNR at my house, and therefore improve my throughput.