DAWN VS django-rest-framework-gis

It is very important to get up to speed with the basics by going through this website first and foremost: https://github.com/berlin-open-wireless-lab/DAWN

By the way, which degree of configuration did you set up? https://github.com/berlin-open-wireless-lab/DAWN/blob/master/CONFIGURE.md

It's not out-of-the-box perfect, but I've had decent luck using DAWN, which targets openwrt, to get decent bandsteering. I can always ssh in (I personally havent been interested in installing/trying the "luci" web interface) & move someone between bands if I need to.

Also, this setup works across & steers clients between my multiple access points!

It's amazing being able to ssh in and see a map of what signals each AP sees. DAWN periodically asks clients to help map, so even if the AP's are on different bands, you can still compare what the signal would be if the node moved.

We'ee finally living in a pretty good time for open source wifi. A pity how only a couple chips have support (select MediaTek and Qualcomm) but wow things have gotten much better.

https://github.com/berlin-open-wireless-lab/DAWN

I'm a long term openwrt user, finally even set up DAWN[1] to orchestrate multiple APs (sometimes a win, sometimes worse than manually picking).

I really hate to say it, but even for the geeks, I'd generally recommend some kind of mesh product. A roommate installed a parallel Google mesh product & it worked so well, had such great coverage. Google spent hella time making this product good, thanks in part to Google Fiber. There's a great talk on their finding from Avery Penarunn back at the first netdevconf[2]. I'm sure many of the competing products are also very very capable. They have a level of reliable, just works intelligence, even over wifi mesh, that makes me envious.

There's sadly little new news in the OpenWRT space. There's some on-the-cusp work with some new Qualcomm chips that might start working in mere weeks or months. But we've been running the same wifi4 & wifi5 hardware for half a decade now, basically, with no real options. The Netgear X4S R7800 is still the go-to openwrt router. MediaTek has some wifi6 capable routers[3] targeted to the low end, which are nice & available & cheap, which is great, but decidedly low end.

[1] https://github.com/berlin-open-wireless-lab/DAWN

[2] https://www.youtube.com/watch?v=yZcHbD84j5Y

[3] https://openwrt.org/toh/views/toh_available_16128_ax-wifi

For central controller management there is DAWN (https://github.com/berlin-open-wireless-lab/DAWN) but I have not used it yet so I can't tell you how well it works.

A quick search brought up https://github.com/berlin-open-wireless-lab/DAWN on the first glance it looks more like hackers-hand-on than run-and-deploy :D

AFAIK this is what DAWN is for: https://github.com/berlin-open-wireless-lab/DAWN

Is this something similar to OpenWISP? It all sounds cool, but might be an overkill for small installations…

[0] https://openwisp.org/

Haven't tried it yet, but I think OpenWISP is what you want.

or https://openwisp.org/

All my searches are pointing to OpenWISP

The only "single pane of glass" open source solution I've found like this is OpenWISP. It works along with OpenWR based devices.

I guess if you want to see what is out there, take a look and openwrt and openwisp

OpenWRT is missing a big piece of the puzzle: configuration management and the ability to work with a "controller". OpenWRT is currently great at running stand-alone but has essentially zero support for being part of a "fleet" of devices managed centrally.

This means something as simple as changing the network name or password requires changing it on every single access point manually, and even worse if your mesh system relies on sharing frequently-changing state between devices.

OpenWISP tries to address this problem: https://openwisp.org - I suggest you check it out and solve the configuration management problem first.

The actual "mesh" part is actually relatively easy. Most commercial systems use basic Linux networking tools, HostAPd (sometimes with custom improvements, but this all ends up upstreamed or reimplemented upstream given enough time) and custom glue code to tie them together. A "mesh" system is typically a user-facing network being broadcast by all APs (with shared settings such as name and password) and an invisible, "backhaul" network each AP hosts (either on a separate interface or on the same interface as the AP - I believe some wireless cards can act both as AP and station as long as the channel is the same) and the other in the path connects to, and the glue code handles configuring all of that. 802.11s is also an option that can be used, and I'm pretty sure all of this is already possible to configure manually in Linux - what's lacking is the "glue code" to set up & manage all of this automatically.

In your project did you end up deploying something like django-rest-framework-gis? I have found great results with it. Mainly using PSQL as the backend. I found that the built in Django GeoJSON Serializer can become a little slow with polys like land parcels but it will get the job done and if you can get way without deploying DRF then it maybe worth the trade off.