Wireless Access

Hi!

 

I´m using IAP-275s for mesh on the .ac radio with 4.2.0.0 code. The deployment is inside a physical tunnel construction and we´re trying to reach the drill in the front with wireless coverage. The last bit will be a mesh hop to a mobile IAP-275 which will provide coverage to the drill.

 

My question:

Will the channelwidth used on the 5ghz radios affect the maximum range I can expect from the last mesh hop in any way? I´m thinking it shouldn´t but I want to double check. The throughput needed is very low.

 

Cheers,

Let's Ask Eric @Ejohnson :-) He is the man!

Thanx Sean! See you WLPC_EU by any chance?

What you set shouldn't have any major impact on range, the radios will negotiate MCS based on SNR, so it's more about channel availability and your throughput requirements (if you only need 10-20Mbps, might as well just use HT20 if the distance is within). How far will this link be?

 

Look at the Rx sensitivities on the datasheet. The higher MCS rates in HT40/VHT80 have much lower receive sensitivities (as well as higher dBm power output). Such as it is, lower MCS rates offer better Rx sensitivities and higher power. You can still set something higher if the spectrum is clear, but it still may negotiate to something like MCS7 anyway. The point is, why set for HT40/VHT80 if you know from the range that it won't make that threshold anyway.

The link will be up to 500m before a new power cabinet is put up with a fiber connected AP. According to my calculations and assumptions that a tunnel is quite good for focusing the RF I hope we will be able to get a few Mbit/s through on that distance with the internal antennas in the IAP-275.

 

Down in the tunnel there´s nothing coming through from the outside so the channels are all available. It will probably be alot of multipath going on though.

ok

 

To answer this we first have to understand the concept of integrated noise in the band.

 

To determine thermal noise power we use the expression kTBF

k is Boltzmann's constant

T is temperature in Kelvin (usually assumed to be 290 K (17C))

B is the channel bandwidth

F is the noise figure

 

to put this is dBm we take 10*log(kTBF)  or 10*log(kTB) + F(in dB)

 

Since we are only worried about bandwidth we can see that doubling the bandwidth will increase the thermal noise in the receiver by 10*log(2) = 3 dB

 

So from 20 to 40 is 3 dB

from 20 to 80 is 6 dB

 

If 20 MHz is 100% of the range

40 MHz is 71% of the range

80 MHz is 50% of the range

 

160 MHz is more interesting

in 160 MHz the radio is actually split into 80 + 80 MHz. So each 80 MHz block has the same integrated noise. HOWEVER, in 160 MHz the antennas are divided across the two block so that reduces the MRC (Maximal Ratio Combining) beneift by 10*log(2) or 3 dB more. (i.e. a 80 MHz 4x4 radio becomes 80+80 2x2 radio)

 

so relative to 20 MHz 160 MHz is 9 dB less or 35% of the range of the 20 MHz channel with the full antenna set.

 

So for a system running MIMO and the same MCS rate is considered

20   MHz 100%

40   MHz   70%

80   MHz   50%

160 MHz   35%

 

Hope this makes sense and answers your questions.

@Ejohnson and another reason the range is better on the 20 Mhz is because it performs much better over a multipath delay ?

Multipath is related to the long and short guard interval. As these are the same for all channel widths and the OFDM tone size is the same the multipath tolerance is unaffected.

Thanks

And that, ladies and gentlemen, is a perfect answer :) I will turn this setup down to user 20Mhz channels.

 

Big thanks Ejohnson!

 

Cheers,

20 Mhz channels have lower
throughput but they have longer range and since your throughput requirement is low you should consider using 20 Mhz vs 40 Mhz

Thanks Victor, any chance you could elaborate why 20mhz channels have longer range than 40mhz or 80mhz on 802.11ac? Harder for Rx radio to decode wider channels?

All of your deeper questions will be answered in Eric and Peter's 802.11ac deep dive.

 

http://community.arubanetworks.com/aruba/attachments/aruba/Airheads-Conference-Vegas-2015/15/1/802.11ac%20Wave%202%20technology%20deep%20dive%20and%20deployment%20recommendations.pdf