A curiosity question: If a node has its bandwidth set to 5 Mhz, are the transmissions centered on the set frequency with energy distributed similar to double-sideband transmission? My apologies if this question was already answered elsewhere.
Theme by Danetsoft and Danang Probo Sayekti inspired by Maksimer
Joe AE6XE
Apologies for the confusion. Your first statement explained what I was wondering about. You are correct about duplication of information in the case of double sideband. I was attempting to show "consumed bandwidth" as a function of the modulation. Obviously, 802.11 modulation is different. So, in theory, I could choose a center frequency which is 5 MHz from the band edge and run 10 MHz data bandwidth without the transmission exceeding the band edge.
Thank you Joe.
Yea, that's right. Although for other's reading this, as I'm sure you know, there will always be some limited emissions out past the channel edge. (opportunity to plug another forum post :) .)
The part 15 rules these devices have been certified for have some specification about the max power emissions allowed outside the 20Mhz channel. The non-802.11 systems can see interference in close proximity. We have seen in SoCal issues with our ATV cousins at a couple of Mt top sites. When they have a receiver in close proximity on the tower, within about ~75Mhz of a mesh node, they can detect some interference in 5Ghz.
To get an idea of how much energy and how clean the waveform is take a look at this post and the charts. https://www.arednmesh.org/content/spectrum-charts-and-take-aways . Our ability to cut channel bandwidth in half and keep max power settings can increase this out-of-channel power -- as shown in these charts 'on the bench' . But, keeping in perspective this is relatively very low power close to the background noise.
Joe AE6XE
Taking into account the extra "noise" from a well mannered data channel, sounds like a good idea would be to choose an appropriate center frequency which would give a gap. Example: 10 MHz data on a center frequency positioned 10 MHz from the band edge. Thus giving a 5 MHz gap between the intended data edge and the band edge. Just a thought experiment
Again, Thank you Joe.
The 2.4Ghz band (and 900Mhz devices) can use this mode for beacon and other broadcast packets -- lowest rate for compatibility with these old devices. On a mesh node look at the following file (nightly builds, path will be a little different for older firmware versions). The statistics show how much a given rate and modulation method is being used:
/sys/kernel/debug/ieee80211/phy0/netdev:wlan0/stations/<MAC address of neighbor>/rc_stats
"if" there is a 'mode' in the output of 'CCK', this means 802.11b is in use with DSSS and 22Mhz. You'll not see this in 5Ghz devices as the standards are all OFDM.
The outer edge carrier waves (of the 64 total) are considered part of the modulation and play into guard gap between channels and I think also help to minimize the peak to average power range of the signal -- an issue to keep in check to not go out of the Power Amp range. Some charts don't count these carrier waves.
A guard gap is always good, if the luxury exists. Here in my area we are doing even channel #s in 5Ghz with 10Mhz channel slots for everyone to align into and minimize possible interference while having the maximum channel options between different groups.
There have been tests in the past to see if the throughput of a node could be affected by another node on the same tower on an adjacent channel. Didn't find an impact when both were loaded down streaming data vs the one node. I believe this was on 5Mhz BW on ch -2 and ch -1, Andre might confirm.