I am new to this group. I do have a network background and have setup building to building wireless networks in a past life. Attached is a high level diagram (Future-Mesh.pdf). This is what I am proposing and the only high ground mesh node that is approved so far is in Brunswick, Maine (direct approval from Bill K1MNW and placed on the top of his 250 foot tower). I have a home mesh node up and running at 60 feet and above tree line. The first mesh node with great coverage will be going up at K1MNW's QTH (in the lower left of the PDF). Before I get to far down the road I just wanted to get some rules/guidelines in place so when future nodes get added things will continue to work. In my design all nodes will be at 5 Mhz bandwidth and channels that are used at each site should not overlap. Knowing that the FCC is taking away the 3Ghz band my focus is around 2.4Ghz and 5.8 Ghz. Also knowing that changes are coming to 5.8 Ghz I thought it would be wise to focus on using channels 179 and above. Im my case I picked channels 180, 182 and 184 to minimize any adjacent node interference.
Please take a look at the attached PDF and I would appreciate any feedback you can provide.(other channels perhaps?, any other channel in the lower part of the band that would also work? Anyone have some guidelines you might already have?)
Thank you 73 Bill NG1P
Please take a look at the attached PDF and I would appreciate any feedback you can provide.(other channels perhaps?, any other channel in the lower part of the band that would also work? Anyone have some guidelines you might already have?)
Thank you 73 Bill NG1P
File Attachment:
Hi Bill. This is a good design. I would offer the following comments:
Andre
Thanks for your thoughts. I should have talked about was I was thinking in the PDF.
- The Brunswick node will support multiple other mesh node connections via the 13 dBi omni. The Oak Hill Wales site will support multiple connections via the 120 deg 19 dBi and the Streaked Mnt node will do the same (trying to save some money). I am using the Nanostation M5 at my house and I assume most home users will do the same if they can.
- My Note in the upper right corner of the PDF was to reference the need to have a vlan capable switch at node locations that have more then one node to connect back over ethernet via VLAN2 Device to Device (DtD) and having a LAN port configured for each node for local troubleshooting via a directly connected laptop perhaps. None of the pictured nodes in the PDF will have a WAN port connected. My home node will and perhaps 1 other.
- Looking at the channels I want to use 180, 182 and 184. I wonder if any of the gear and or antennas will even work up that high? Testing with the nanostation and Rocket at home but not at any great distance (yet). Most antennas only claim 5.9 or just 5.85 for the high end. If this is true I may need to change to other channels? Any thoughts on channels I should pick that would not be impacted by an up coming FCC change or have minimal interference.
Thanks again! Bill NG1P
Hi, Bill:
I did a test of recording the SNR between 5 GHz devices at maximum power and found a (smooth) 10 dB roll-off from channel 169 to 184.
IIRC the test units were Mikrotik LHG and SXT.
The SNR was flat at reduced power (IIRC 10dBm).
I have had very good results using the part 15 channels (132-165) when otherwise unoccupied.
With a pair of LHG-5HP-XL, I am getting a TxMbps of 100+ on a 13.5 mile link on channel 140 (20 MHz BW), SNR is 30+dB.
If you feel compelled to use ham radio only channels, consider staying closer to the manufacturer's specifications.
Wi-Fi scanning locally typically shows (part 15?) devices on channels 132 or 136 and 149 to 165.
In practice, in remote areas, on PtP narrow beamwidth links,
I do a 20 MHz Wi-Fi scan and test the channels 132-140 and 145-169.
I recommend avoiding 5 MHz bandwidth unless necessary to achieve a working SNR.
YMMV,
Chuck
Hello Bill,
My home QTH is in Phillips and I would like to participate with you in building the network here in Maine. What would be the best gear for me to get that would be compatible with what you are using?
John WA1KLI
Met you in Augusta at the hamfest in Feb. For right now I am focused on using the 5.8 equipment and just in the Bunswick/Topsham area. With interest and support I would like to get more coverage (still pending) I should have the high ground node up and running soon maybe this weekend. This will provide coverage down to Portland. Should be fun.
73 BIll NG1P
I have sent you two emails one about a month ago on this topic and another one today with more info. I wonder if you got them?
Bill NG1P
Hello Bill,
What with Covid-19 lockdown this past year this slipped through the cracks. Not sure what happened to your emails, I appologise.
I would like to participate, what is the current status of AREDN here in Maine?
Thanks,
John WA1KLI
1) 802.11n specification builds into the 64 carrier waves fitting into the channel mitigation for both inter-symbol (inter-carrier_wave) and inter-channel interference. The carrier waves on the outside edges are modulated accordingly as built in guard gap already. I use 170, 172, 174, 176, 178 all on the same tower on 10Mhz channels -- major site overlooking Orange county, Los Angeles county, Riverside, and San Bernadino county. No issue with interference between nodes -- one is the 40 mi link in #2.
2) For the distances you have, there should be sufficient SNR to obtain double the link rates and through put by going to 10MHz channel widths. We usually see max MCS15 rates on 40 mile links on 5Ghz and 3GHz with 30dBi rocketdishes -- 65Mbps. At 5MHz channels the max rate is cut in half and max 32.5Mbps.
3) I don't think I'll ever install a 2GHz coverage in the future, unless away from a city. Only 1 usable channel at 10MHz width. My experience is 2GHz has the most noise and performs poorly in comparison to 3 and 5GHz.
Joe AE6XE
73 Bill NG1P