Our tunnel mesh island seems to be working, but I am trying to understand the RF parts of our local mesh network.
Our local mesh has five “hub” nodes. Our hubs are collections of hardware within about a 3-mi radius on tall buildings or water towers – each with one omni antenna or three sector antennas with Rocket radios - supporting maybe ten total active user nodes. This morning, one directional node of one of the hubs had eleven Current Neighbors in Mesh Status: six show RF but none is connected (one is blocked due to dtd with a co-located hub sector node, five are blocked due to signal or latency), four nodes show (tunnel, active) status, and one shows (dtd, active) status.
One of the blocked RF links is to distant hub. The Mesh Status display for this pair of nodes shows no LQ, no NLQ, and no Distance, but Quality = 29%. The SNR shows 30 dB and the physical distance between these two hubs is 1.2 miles (per Google Maps). The status shows (blocked, latency).
What could be keeping these two hub nodes from communicating? How can there be a latency issue with two nodes that are about 6 microseconds apart? Is this an artifact of 3.24.4.0?
Vy 73,
Tim K5RA
Hope this helps.
Orv W6BI
Orv -
Thanks for the prompt reply.
The AREDN docs on Mesh Status Display indicate that there is a distinct status for (blocked, retries). Can too many retries also generate (blocked, latency)? How to troubleshoot a system if the indicators are misleading?
Here is an example closer to home. I have two Rocket radios, one at 77 ft with an omni antenna, the other about 15 ft AGL inside a building with a sector antenna looking out a window. When I could access it, the radio with the omni has never shown much connectivity - maybe 4-5 stations. The Mesh Status display of the radio inside sees the radio outside with an SNR of 34. The inside Mesh Status display shows no LQ, no NLQ, no distance, and the Quality is 0%. The two nodes are less than 200 ft apart. The status report is (blocked, latency).
The outside radio has an SNR of 19 and a Quality of 22% in a hub node about 6.3 miles away.
Thanks.
-Tim K5RA
I read your post and cringed when I read "omni antenna".
Then, again, when I read about an single hub node with "ten total active user nodes".
Then, again, when I read about
"one directional node of one of the hubs had eleven Current Neighbors in Mesh Status".
Did I read that correctly?
There is a channel with 12 active nodes?
"What could be keeping these two hub nodes from communicating?"
Too many stations on the same channel.
Perhaps, in ad hoc mode, nodes are never 'connected'...like when AX.25 are connected.
Nodes are kinda aware of neighbors.
"I understand there could be hidden nodes, It only takes 2 nodes to have a a hidden node. :-|
It takes a few as 3 nodes for a hidden node issue to cause communication to seriously degrade.
I see already replied about:
"our hub node of interest" and " duplicate (rf and dtd)"
Co-located and DtD'ed nodes are fine, but is this 2 co-located nodes on the same channel? !!!
I am beings facetious, but how can 2 co-located nodes on the same channel not be exposed?
How can there not be hidden nodes among them.
Perhaps, I do not understand your network.
You mentioned of 'our hub node of interest' seeing other nodes of SNRs, 29, 23/26, 18,13, 10, 9, and another with no entry.
That seems to be 8 nodes on the same RF channel!
Is that correct? This was planned?
73, Chuck
Can you provide a simple drawing that shows your hub locations with what radios / antennas (direction pointed if directional), and channel / BW. Also some idea how many end user stations are expected to be attempting to access each of the hub stations.
I'm with the other guys that you have created a massively congested network and are seeing the issues from hidden nodes. retries, etc.
Glad to hear the attachment worked.
Here is yesterday's work as txt file attachment. The mesh is on 2.4 GHz, Ch -2, 5 MHz.
--Tim K5RA
Nice map.
Thank you.
It explains much.
Reply to your .txt attached.
Our local network at 2 and 5 GHz attached.
73, Chuck
Orv W6BI
when doing a 20 MHZ BW channel scan, the highest channel used around here is Cox Communications on channel 165, so they'd be using 163 164 165 166 167 ?
Using a one channel buffer, the following would be recommend as an ideal channel plan for 10 MHZ BW ?
184 183 182 channel 183
181 buffer
180 179 178 channel 179
177 buffer
176 175 174 channel 175
173 buffer
172 171 170 channel 171
169 buffer
168 buffer
Richard ko0ooo
Channel overlap may cause interference.
73, Chuck
A 20 MHz bandwidth Wi-Fi 802.11gn signal on channel 165 would be using
the upper half of channel 163,
all of channel 164, 165, 166, and
the lower half of channel 167.
A 20 MHz bandwidth AREDN signal on channel 169 would be using
the upper half of channel 167,
all of channel 168, 169, 170, and
the lower half of channel 171.
73, Chuck
Joe also says:
"Another thing to look at is that often times it comes up with what about adjacent channels,
signals on an adjacent channel, the signal spectrum of twenty megahertz signal...
These align right on top of one another side by side in the channels and are designed so
that to work and to be optimal and
you shouldn't expect interference with two twenty MHz signals right next to each other."
In my testing I have found this to be true.
I have operated multiple co-located 5 GHz nodes at 20 Megahertz bandwidth
on adjacent channels without detectable interference.
The lower shared channel numbers are no problem. You only need to avoid other devices that have strong signal strength and if you see the SSID of other users and choose to avoid as a courtesy, that is fine. For example, when choosing a channel at a hospital that lets us be on their roof I avoid the channels I see they are running their wifi on ... don't bite the hand that gave you roof access!
If another user has low signal on 140, that does not mean totally avoid 139-141 ... this stuff coexists pretty well. A lot of the signals you see on lower frequencies are home wifi routers and your SSID won't be picked up on them and you can't overcome their signal strength from the viewpoint of inside their house ... unless it is super strong signal as seen by your device and they are your next door neighbor. The worst that would happen is degredation of link quality and speed if you park a 150 right next to another 150.
Ed
Ed