Turning off Radar Detection on DFS channels?

Yea, we're going in logic circles. I saw the EU law you highlighted - and we already discussed:

Preventing install of software like OpenWrt and Libre CMC is how manufacturers do that til this day. That's been evidenced and proven thru the many post-2015 devices that now support OpenWrt in which those features must be circumvented so the end users can have their device "accept software". Nonetheless, it seems you think it means something different - and I can respect that. But without a reference, it seems you're just making a blind assertion.

I agree to disagree, thanks for the hearty discussion.

2 Likes

That is no proof because OpenWRT contains the non-free blobs for the wifi chipset, only LibreCMC does not. It is the very reason for the existence of LibreCMC.

See here:

The OpenWRT/LEDE project does help give users more freedom than they would otherwise have, but some of the supported devices and packages are non-free.

If you think I am making a blind assertion, can you please name me one other way for a manufacturer to ensure the wifi chipset is "not accepting software and/or firmware which results in the equipment no longer being compliant with the DFS requirements" other then checking if the loaded software (the one loaded onto the wifi chip by OpenWRT) is actually the approved one? How can the manufacturer make sure no software that does not comply with the DFS rules is loaded? There is no other way for the manufacturer to comply with this law, at least no way I can see.

I've read that post by Naftali you linked twice (in fact, I read it on 17 Nov 2020 when it was posted). But as I noted it seems like a plug for LibreCMC over OpenWrt. I don't get the point.

At least in FCC land, manufacturers are not required to do what you asked (I even provided a quote from the ET). I'm not sure how anyone can reasonably expect such a thing.

Despite this and regarding the EU law, I already explained how manufacturers do this:

  • they prevent serial access by not soldering header pins
  • they configure the web GUI to prevent consumers from uploading unapproved software - as you correctly noted, it checks to see if the software is from the manufacturer

Those are 2 examples, I hope this clears up the confusion. I understand this is not as hack-proof as you'd like, but it's acceptable.

The other alternative implies OpenWrt is in some conspiracy to quietly switch blobs and the developers simply won't say so, that seems far-fetched.

1 Like

That would be locking down the device.

So they can either lock down the the whole firmware or only the wifi part of it. But I meant how can they do so without locking the device or the wifi chipset down? There is no way.

Correct, that's the requirement (i.e. that the final end-product be locked down). That's the manufactured end-product (i.e. the assembled device needing the FCCID, inspection/test and approval). I think you're making a differentiation that the radio regulations don't.

They're not required to lock down a chip that has to be soldiered to a board, modular, etc. Keep in mind, the board needs approval not the chip. I provided the FCC example of that:

https://www.ecfr.gov/current/title-47/chapter-I/subchapter-A/part-15/subpart-E/section-15.407

Here are the regulations from FCC land (because you find such laws across the globe):

(i) Device Security. All U–NII devices must contain security features to protect against modification of software by unauthorized parties.

(1) Manufacturers must implement security features in any digitally modulated devices capable of operating in any of the U–NII bands, so that third parties are not able to reprogram the device to operate outside the parameters for which the device was certified. The software must prevent the user from operating the transmitter with operating frequencies, output power, modulation types or other radio frequency parameters outside those that were approved for the device. Manufacturers may use means including, but not limited to the use of a private network that allows only authenticated users to download software, electronic signatures in software or coding in hardware that is decoded by software to verify that new software can be legally loaded into a device to meet these requirements and must describe the methods in their application for equipment authorization.

(2) Manufacturers must take steps to ensure that DFS functionality cannot be disabled by the operator of the U–NII device.

How would they be able to do that without locking either the device (OpenWRT cannot be used) or the wifi chip (OpenWRT can be used, but only wit the non-free software from the manufacturer) down, so only the approved non-free software from the manufacturers of the device or its parts can be used?

Information already provided; but I'll provide the quote again. I think it's fair that we should end the discussion and agree to disagree:

To be clear - I meant acceptable by radio regulators like the FCC and ETSI.

I understand you're asserting there's more complex lock down procedure, and your grand speculation on the WiFi blobs proves that.

Nonetheless, manufacturers of assembled WiFi devices have been doing the above since 2015 - and radio regulators worldwide have been OK with this this and approving such devices.

Lastly to be clear you're making a distinction where one exists, the chip manufacturer isn't considered the manufacture of the completed and assembled device (i.e. what we're discussing) - so I'm lost as to why you assert (without any reference) their blobs needs to be locked down as claimed in your comments.

2 Likes

@Emanuel

This Wired article will make it more clear, from 2016 (took a while to find it):

1 Like

Oh, I see your point now. You mean the way the FCC and other regulators make sure wifi routers cannot be operated in a way that would violate the laws is much laxer than I thought.
That is possible as well.

I found the non-free ath10k binary blobs with their license and a Wikipedia List of wifi firmware and divers. As you can see ath9k does not need firmware for the wifi chip as the wifi chip does not have its own CPU. With ath10k and ath11k, the wifi chipset does have its own CPU.

From the article you linked:

“Routers are built around a System on Chip, with the CPU and radio in a single package,” Hackaday’s Brian Benchoff explains. “The easiest way to prevent modification of the radio module would be to prevent modification to the entire router. Some would call it fear mongering, but there was an expectation these proposed FCC rules would inevitably lead to wireless routers being completely locked down.”

It just seems strange that the newer chipsets have their own CPU, but you are right, I cannot know.
No firmware for WiFi 5, WiFi 6 or WiFi 7 chipsets has been reverse engineered yet and I can find no such projects, which I just find a bit strange. Does nobody care about that any more or is there something else hindering development? I don't know. I looked inside the (permissive) open source licenses of some parts of the non-free firmware. It includes much cryptography software, seems like the non-free firmware is responsible for the security of the wireless network. I also found licenses of C files for checksums. So the crypto software would definitely be there. If it really does what I thought the law requires, I cannot know. Might as well depend on how much Qualcomm might dislike reverse engineering.

I first thought that it works this way because I know Intel and AMD do that with their processors. Firmware is digitally signed by Intel or AMD and the processor refuses to run on a firmware not signed with the private key corresponding to the burnt-in public key. (You can flash it, but it will refuse to run.) The firmware I am referring to is the one running on a processer separate from the main processor best known for the Intel Management Engine which is a part of it. Some speculate that the Intel ME and AMD PSP might have been influenced by the FBI:

In the context of criticism of the Intel ME and AMD Secure Technology it has been pointed out that the National Security Agency (NSA) budget request for 2013 contained a Sigint Enabling Project with the goal to "Insert vulnerabilities into commercial encryption systems, IT systems, …" and it has been conjectured that Intel ME and AMD Secure Technology might be part of that program.[78][79]

It is also known that the US government had a secret way to disable the ME on their computers. Looks like something is not right.

And I know about the planned Microsoft Pluton which is the next step in digital restrictions management (DRM) after TPM2.0 and might also be used by corporations and intelligence services aiming to hide something from the public to arm themselves against whistleblowers (although it is not clear yet how far Microsoft and AMD will dare to go, it will depend on the reactions, though they certainly test their limits every few years).

Just so you can see where these thoughts are coming from. :slight_smile:

Perhaps we can agree that we do not know how the laws are currently interpreted by the manufacturers? Otherwise, let us agree to disagree.

1 Like

Intentional (?) engineering blunders of history:

Put 1 watt wifi routers and 1000 watt microwaves on the same frequency in the same domain.

Solution:

Now add microwave weather radar in the same frequencies. Ooops already did that - in use since the 1950's?

Whoever designed this should be fired. Problem solved. If this solves your problem please mark this topic as solved.

In 2023, the private American company Tomorrow.io launched a Ka-band space-based radar for weather observation and forecasting.[13][14]

Problem solved. I'm 2 for 2.

2 Likes

There is a difference!
A micowave owen is a screened faradays cage normally connected to earth potential so it isn’t a transmitter.

The accesspoint is a designed transmitter.

And pulse-doppler radar is more of a 70-80’s thing.

And 5GHz wifi is more of a 2010+ thing.

4 Likes

Also, the microwave oven has priority on the band. The WiFi is actually the unlicensed device.

3 Likes

This is actually the actual problem. Wifi has become so big in the world that it would probably work better if it got a dedicated band only for wifi.

The 5GHz and also 2,4GHz band now in use is free for all to use band so the problem is pretty much guaranteed.

1 Like

You forgot the frequency at which water boils (and not something else in our bodies nearby - like DNA) ?

And will bounce a radio wave [effectively]?

And will hence naturally stop propagating in our atmosphere?

Sorry, you must complain to Nature's God. :pray: :prayer_beads: :palms_up_together:

It's Physics. :teacher:

1 Like

Yup. https://en.wikipedia.org/wiki/13-centimeter_band

Microwave (ISM) applications are the primary user. Secondary and tertiary users must accept interference from the primary user(s). They must also not create interference for the primary user (fortunately, this is not really something possible for a microwave oven...)

Amateur (ham) radio usage in compliance with regulations (identification, no encryption allowed unless your keys are widely published somewhere) is secondary

Non-ham unlicensed users are tertiary. They must accept interference from primary and secondary users, and must not interfere with primary and secondary users.

For the U-NII bands, primary users are radar, unlicensed devices are secondary - they must accept interference from the primary user, and must not cause interference to the primary. Unlicensed devices interfering significantly with TDWR is well documented - https://its.ntia.gov/umbraco/surface/download/publication?reportNumber=11-473.pdf

As others have pointed out, while TDWR is the most common thing DFS is designed to detect and avoid, military radars are also a possibility. I suspect this is why DFS is hard-locked-out where I live despite no known TDWR station for miles. I'm in the process of investigating why my company's network sees what appear to be false positive DFS trips 3-4 times a day, and actually working on going the other direction from OP (forcing DFS to trip) in order to test how our products handle an AP changing channel.

3 Likes

I think it would always be possible to select Russia or India as country,
then you should be able use channels 52 till 68 without DFS being applied, not sure if OpenWrt check this on country or not.

On the other hand, why would you try to turn this off?
If you are certain not radar is nearby, you will never have to change channels. and if a radar is nearby this will help you get out the way for critical information.

If an ambulance passes you on the street, do you go out of the way, or are you stubborn I make sure he can't pass you?

2 Likes

Well that's kinda the point of 6Ghz, its not entirely a frequency to WiFi alone but it specifically checks if any licensed frequency is used nearby and locks those out, rather than the false-positive prone DFS.

Of course this will upset OP even more, as 6Ghz is even more user restricting as it requires polling on an online database before it will work at all in order to check which frequencies are licensed in your area.

1 Like

I am wondering (as simple as that) who will pay the bill - if someone ever get back to you of using illegal channels and if you ever get a bill:
According to the above conversation it seems that OpenWrt is liable for any violence of FCC or EU regulations or the manufacturer - but are we sure? Did it ever happen? EU is promoting the end users to able to choose their end user equipment and stop these ZTE/Huawei- etc end-devices as mandatory by ISP.
At the end it is your choice, and you should be liable for any plane falling down - because you used DFS channel.
At least it is my opinion.
Kr.
K

Unless one lives in those countries, that too violates the law. I think that kinda was the point of the original inquiry, until the OP was informed DFS too was the rule.

  • Linux is compliant - not sure how OpenWrt came to mind
  • Therefore nothing has changed (this is even more so in the example of every device already running a OEM fork of OpenWrt already, or Linux rather)
  • More importantly nothing changed - because we agree the code on the chip is unchanged
  • Even more it's the manufacturer's blob
  • I was lost how the entity who didn't touch the device - ended up in your conclusion responsible for something
  • :bulb: That's why nobody suggested (or shouldn't anymore) doing anything to become not "compliance with DFS requirements" (EU) nor changing anything that would "ensure that the devices will operate as authorized by the Commission, thus reducing the potential for harmful interference to authorized users." (FCC)

Now you see why discussing altering the behavior is really a no-no. I know that it just feels like someone has made some complex rule - but it just isn't so.

As I noted:

This continues to be a long discussion for those just learning a bit about regulations.
So what is the end user usually not allowed to do?

Another FCC land document:

The purpose of Section 15.203 is to prevent attaching any other antenna(s) [other than one(s)
approved with the device] to a Part 15 transmitter. All antennas for use with the approved device
must be listed in the application.

~From a document entitled "BASIC EQUIPMENT AUTHORIZATION GUIDANCE FOR
ANTENNAS USED WITH PART 15 INTENTIONAL RADIATORS"

BTW - around that time, manufacturers started redesigning devices that weren't meant for outdoor use (give me some time, and I'm sure I can get an article on that too). As most are aware, the antennas of today's indoor APs are generally integrated as a part of the whole design, non-removable and would be difficult to integrate another [well] without the skills necessary to do so. If there are detachable antennas, the models are also listed and approved by the FCC. Que the history of the HGA7T high gain antennas for the old WRT54G's that had [removable] antenna connections.

there is a clear difference between rogue hackers and those who prefer and support open source projects. Here you will find very few if no one willing to offer you the advice you're looking for.

2 Likes