America's crackdown on open-source Wi-Fi router firmware – THE TRUTH America's broadband watchdog is suffering a backlash over plans to control software updates to Wi-Fi routers, smartphones, and even laptops. In a proposed update [PDF] to the regulator's rules over radiofrequency equipment, the Federal Communications Commission (FCC) would oblige manufacturers to "specify which parties will be …
I'm astonished, how on earth could they implement or even apply these rules?
Phones/Routers etc.. as we have found recently need and should have the ability to be updated, if this is applied then they will be locked down from the start so a whole hatstand of vulns would never be corrected.
or maybe (Tinfoil hat time) that's what they want?
A long time ago, when men were men and they hacked into satellite TV systems, there was a certain bit of hardware often jammed into the Smartcard slot. A long gangly PCB with wires and things hanging out. As it happens, the circuit emitted spurious signals on the emergency beacon frequency of 121.5 MHz.
Such signals will eventually attract a large helicopter. Such search and rescue helicopters are obviously equipped with precisely the sort of highly sensitive beacon receivers and DF/Homing equipment needed to place the noisy thing hovering directly over one's house at about 11 PM.
Just a bit of history.
As it happens, the circuit emitted spurious signals on the emergency beacon frequency of 121.5 MHz
Only if it had been made by a blind, tetraplegic colour blind hedgehog whilst in a bag.
Other than that it emitted no such RF or harmonics.
Or at least the few hundred or so I made didn't!!!
Whilst at Uni I had a house with a dish on the wall and I thought I'd get a receiver and see what was available free to air. For the record I was not doing anything related to Star Trek or any of the Seasons of the show, not that I wasn’t tempted for educational purposes. I tuned into VH1 UK and found the picture scrambled but VH1 Germany on the same satellite was FTA. A mate came round and saw the fact that I had free music television, Eurosport and a few other channels that were worth watching. So he went home and having borrowed a ladder from the uni hooked up a dish and the Sky Videocrypt Integrated Receiver and Decoder (IRD) a hand me down from his dad after he'd got a new one. On this box however both VH1 channels were scrambled (although the Germany one was clear for about a second) and he was perplexed as to why mine worked. I went round and it was only that channel that didn't work all the other FTA channels were watchable. So I called directory enquiries got the number for MTV Networks Europe called them and asked.
Nice lady says she doesn't know about that but gives me the name of the person I need to speak to who does. The bloke I ended up speaking to says that they have flagged the German version as encrypted so that the IRD scrambles the picture. This is down to the rights being for different countries and the adverts etc. So my mate unpluggs his IRD and with his driving licence + other ID buggers off to Cash Converters and sells it. He buys a bog standard analogue receiver from the same shop with the money and then had free music television like me. I never got round to installing a motor on my dish (did have to do some academic work) but I did have a positioner box knocking around.
Good luck to the CBP checking all the items I bring to the USA for FCC approval.
"I don't get the Star Trek reference......."
The original reason (allegedly) for sky being hacked was when they decided to encrypt SkyOne when they were about to show the final season of Star Trek The Next Generation. Someone wanted to watch the final season and supposedly hacked the videocrypt to allow this without needing to be a subscriber. It became known as the Season hack as a result.
'I'm astonished, how on earth could they implement or even apply these rules?'
Why not? The Digital Millennium Copyright Act and other laws already ensure you can't snoop on–or even worse–alter software, so why is it so surprising the same rule is now being applied to hardware? (Seems to me the timing has come about now as this is the first 'plausible' excuse the 'Establishment' has had to raise the issue of control.)
As I perceive it, any such control by The State would be utterly disastrous and it needs to be resisted and fought at all costs. There are many, many reasons for fighting this, most too detailed to mention here but probably the least of which is security and interference.
Clearly, safety regulation is necessary in specific instances such as the possession of radioactive materials, x-ray equipment etc., but banning the altering and tweaking of general electronic equipment is another matter altogether. With respect to spectrum management (non-ionising EMR), effective and workable regulation has long been in place so that mutual interference between services is minimised to acceptable limits. This regulation has worked well for most of the 20th C. without need to lock out the general (technical) public.
Consider this: since about the time of the introduction of the PC in 1980, access to the workings of technology has inevitably been reducing. The original IBM PC came out with full circuit diagrams and BIOS source code (I know I've still got the manuals) but nowadays we can't even get access to the boot information of our PCs let alone circuit diagrams of our domestic PCs and other electronic equipment. Circuit diagrams were once commonplace and the accepted norm, now they're extinct.
Essentially, the citizenry is rapidly losing access to the working and control level of many different forms of technology: from software, PC hardware to chemical system, to pharmaceuticals, one's vehicle, etc., etc. Both industry and governments use excuses such as security, safety, the equipment's too complicated etc., but the real reason is for industry and ultimately The State to have full control over the technology.
Technology is ever-increasingly important to our modern lives, but locking away knowledge of its operation not only deskills the citizenry but it's also a major threat to democracy.
"They" had a Dream - of a New World, Where Every.Single.Device designed is also a Weapon, protecting and, when needed, defending The American Way of Life (and Graft).
Based on your definition of the American Way of Life, America would wind up full of nothing but looters/moochers who never do anything useful or productive. But I suppose that's the ultimate goal of virtually every government on earth.
Surely, Jefferson, Voltaire, and Adam Smith are spinning in their graves.
Given the open-source tinkerers are responsible for providing the core code used by the vast majority of the routers out there (which companies then take, and plonk security-flawed UIs on top of), this would seem to be a very ill-conceived move.
If the tinkerers can't explore the hardware and get things working properly, that is going to bite us one way or another.
Completely agree.
And not just the routers but the user devices too. Most internet enabled TVs (Linux), most phones (AOSP, clue is in the name), embedded systems.
Yes there's a discussion to be had regarding an improved approach to security from OEMs and end users alike, but this policy suggestion from the FCC is not answer. It would put the brakes on the whole process of innovation and stifle the future of the tech sector.
I have resolved my own phone's exposure to unintended use of radio frequencies (thanks to recently disclosed bugs in Android) by flashing a community ROM. The OEM still hasn't moved to patch the flaws, but I'm covered now thanks to the open source movement.
It looks another Pandora's Box is opening. If operating frequencies can be easily modified at the software level, it will lead to too many idiots doing every idiot thing they could think about (just like we see with lasers and drones), and thereby they will hamper the opportunities of those people actually knowing what they're doing and doing it without causing no harm to everyone. Just, improvements in hardware capabilities and software is making idiots capabilities broader too.
Is Software Defined Idiots the next wave, and how could we control them?
Radio hardware that can be arbitrarily reconfigured in software has been the norm for at least a decade - by now, there isn't ANY vaguely current radio equipment left that ISN'T some form of SDR. Many of those have enjoyed open-source support for a very long time now. It's not impossible to misuse them, but usually you really, really have to be trying. Curiously though, the sky hasn't fallen yet. So let's either back off and regulate actual misuse on a case-by-case basis or start banning things like knives, considering how widely they can be misused as well...
It's like drones. As long as most flying hardware was for very interested people, and usually available in specialized channels, at not so low prices, there were usually very little issues. The day every idiot can find one, and prices are low, it will get one and find an idiot way to use it.
People listening to police/firefighters and ATC frequencies have been doing it for a long time. Just nobody ever thought to interfere (but in very rare cases)
The issue is exactly this, how to identify the idiots. Knives are dangerous, but it's easy to identify an idiot branding a knife. And we're told since childhood knives are dangerous. It's worse when people act without thinking about the consequences. Drunk people do kill people driving. Is it driving or drinking bad per se? No it's the combination of both, idiots who can't think about the consequences of what they do. That's why car have a plate, to be identified. Too often, without accountability, some people may too easily become irresponsible..
"The day every idiot can find one, and prices are low, it will get one and find an idiot way to use it."
This assumes there are no rich idiots.
If something can be misused, it will be, no matter, the affordability, availability or legality, those are just hurdles which some individuals find easier to leap over than others (or merely "trip over when they are running around shouting"*)
*Ahh, another Prattchet ref.
"...there isn't ANY vaguely current radio equipment left that ISN'T some form of SDR..."
Software in the IF strip is slowly getting more common, but it's hardly universal.
The rest of the RF circuitry is still almost universally hardware.
Even those $10 USB SDR sticks contain a *hardware* tuner chip in front of the ADC section. They certainly do not sample directly at L-band (~1 GHz area), I can assure you of that.
I am just wondering if this is an actual problem, or just an imaginary one. There are plenty of things in the environment which have actually been observed to interfere with essential radio communications on numerous occasions, like rain. I am just not sure this is one of them.
Yes there is a problem - well from the regulators perspective. From what I've been told, it appears that quite a lot of WiFi enabled devices don’t work very well on the listen before talk function (essential for avoiding other legitimate users of the band like weather radar) and the associated frequency hopping and transmit power control functions. Therefore the FCC are trying to crack down on the matter
Another problem is that (as far as I can see from the proposal) it will only apply to new devices. Those that are out there already - think of how many DSL models / domestic APs that are already out there and gathering dust under a table in the lounge are going to keep on going with very little support from the manufacture.
A partial solution (and not going to solve poor coding / self declaration by the manufacturers) is the regulatory wireless database which attempts to keep to up to date all those SOHO devices (and other peces of kit) with world wide regulatory rules but relies on people knowing it exists and then contributing (and the kit using it properly).
I fear (as the article has pointed out) that FCC policy people have made a poor call on this matter, but like all "unlicensed" spectrum, once the rules have been relaxed, its very hard to regime them in again.
AC due to my employer
From what I've been told, it appears that quite a lot of WiFi enabled devices don’t work very well on the listen before talk function (essential for avoiding other legitimate users of the band like weather radar) and the associated frequency hopping and transmit power control functions. Therefore the FCC are trying to crack down on the matter
It's not quite that simple. Weather radar (the stuff used at airports for safety of inbound aircraft) uses frequencies that fall in three specific channels in the 5GHz ISM (wifi) band (in the US, not sure about other countries). No wifi device is allowed to use the three channels that cover those frequences unless they have dynamic frequency selection (DFS) enabled, which is supposed to monitor continuously for other signals. The problem isn't so much that DFS doesn't work on some devices but that some of them, possibly intended for non-US markets, don't implement it, or don't require it to be enabled in order to select the channels in question.
The trouble is, the FCC has got to try and anticipate problems before they happen; once the genie's out of the bottle it will be hard to put back in.
There's also the potential problem of deliberate misuse of SDR devices to create a denial of service on radio frequencies. We currently have any number of examples of malware that targets home routers for malicious purposes (DNS redirection for example) so there's a real possibility of criminals or foreign agencies deliberately targeting SDR devices.
Those who want to tinker have a definite point, so the mechanism for control has to take this into account; one option could be that only updates from authorised sources can be delivered over the network but anyone with direct access (presumably the owner) can apply an update from a hardware i/o device (e.g. USB).
Yeah, that "software level" adjustment is a real "Pandora's box" of problems.
It makes those home soldering kits, aerials, wires and manual dials and knobs pale into insignificance.
/sarcasm.
So if the old systems were used mainly in the norm by the majority, the same will happen here. Though software allows a quick distribution, it's about the same as a flash mob tricking a community into retuning their two way radios into the local police chatter frequency. Scale of the problem, often scales with the solutions. In this case, it's general virus scanners/software warnings, or general limitations to the hardware.
I remember catching wind of this a while back but when I didn't see the expected coverage hit the major sites I figured it was a bit of FUD. I don't envy the FCC on this topic. I can understand their need to enforce certification, etc but at the expense of BYOF (bring your own firmware).
If only the radio firmware/stack/driver could be abstracted and reside on another piece of flash. Easier said then done on the software side I imagine. That and it would increase the cost of the devices. At any rate perhaps a middle ground could be found where the layers further up the stack were user modifiable without having direct access to the radio itself? Sorry, tired probably not making sense.
Now what are the odd's that companies will start updating their firmware due to these proposed rules? LOL. Let's enforce who can update the firmware! No one ever does....
This could be a reasonable compromise. Allow the thing that talks to the device, and the device's interface to the outside world to be freely publicly available.
Make sure the stuff inside the device does what it's designed to do no matter what. The firmware could be open-source, but perhaps only signed updates by the vendor are accepted by the device.
While not as good as a vetted all-open system, it is better than an all-closed system where you can't even fix things higher up that are either broken or missing. More than once I've telnetted into a router and used the Busybox version of vi on the device to hack a few shell scripts to work around deficiencies in the web UI. (I'm looking at you, Netcomm. There's a couple of NTC-6000 series routers on defence bases that have had such treatment.)
So why ban opensource wifi and not just improper use?
The software stack that runs most consumer kit would evaporate overnight. The industry would then have to spend cash to replace said software stack and the industry would miss out on any innovations that come about as a result of the current system's openness.
There's already a lot of kit out there that is more open, and some of it is built well enough to run for many years.
@Ragequit - I was thinking the same thing.
If the device and its firmware are properly designed, the low-level RF controller should be encapsulated so that is in control of the frequency, and anything that uses it cannot request a frequency outside of the allowed bands. I see no problem in that part of the firmware being regulated, certified, and locked down. Anything above that on the tech stack doesn't need regulation (at least not for this purpose).
The problem with the acceptable bands is that they vary by country. i.e.
https://en.wikipedia.org/wiki/List_of_WLAN_channels#5.C2.A0GHz_.28802.11a.2Fh.2Fj.2Fn.2Fac.29.5B16.5D
So the issue is that effectively the radios can be used in any country and are software selectable for the chosen country. If the firmware allows the country to be set, then setting the AP to Russia gives more available frequencies without that DFS/TPC reducing your signal strength.
They really need a solution would be providing a way for the radios to work out their location and restricting how they operate
i.e. while I can think of how to identify your country if you have Internet access using GeoIP, how do you do it on devices that have limited or no Internet access or incorrect GeoIP details? And while DFS/TPC can help with restricted bands, they rely on detecting an active channel so if a channel is used infrequently you still have the possibility of interference
And this is ignoring any issues with software quality from manufacturers.
They really need a solution would be providing a way for the radios to work out their location and restricting how they operate
… Or the governments can recognise that unlike 50 years ago, devices do regularly cross international boundaries, and that unified worldwide spectrum allocations could save everyone a lot of trouble.
The problem with a unified worldwide spectrum allocation is that either:
a) the allocated range is significantly smaller than what is currently allocated limiting potential uses (check the wiki page for the common frequencies that are unused by all regulatory domains)
b) move or remove existing users to free up space. As a lot of the usage is weather/military radar I suspect the time frame for doing that is measured in decades.
I suppose it would be selfish of me to demand the flexibility of buying a $30 router and making it as capable as a $150 with custom firmware... But I'm relatively poor.
Whatever... All my game consoles are modified, DMCA be damned. I wont be caught mucking about with emergency services coms... So I guess I needn't worry.
Does your modified router transmit wifi signals any differently to the $30 original, or are the changes related to the UI or the way it routes network packets? If you increased the signal strength or transmitted on different frequencies, that would be a problem, but having more control over packet filtering and forwarding should not be a problem.
Anecdotal evidence here...
Locally, someone accidentally wired the live mains into the telephone with through their sky box. No idea how, why or what, but they did. Everyone had no internet for about 8 months until the local utility company finally admitted the entire street being down might be considered a "fault".
So, if something like that can happen, on a massively distributed commonly used piece of hardware, why such a worry about one or two wifi routers?
If this was about cars, and speed limits. We would make the modifications illegal, not put key and lock on every bonnet!
"Everyone had no internet for about 8 months until the local utility company finally admitted the entire street being down might be considered a "fault"."
Two things come to mind here. Firstly - surely mains on a phone line would only affect the subscriber that the line belongs to? The phone lines are designed to be pretty resilient, dealing with surges from lightning and nearby electric lines, plus the POTS ring signal itself is an AC of around 90V. When lightning nuked my Livebox (burnt it out inside), everything "just worked" after plugging in a spare. There's no way an engineer could have been out and fixed the hardware in that short time, so I assume the exchange either failed over to a backup bit of hardware or it absorbed the energy in a less destructive manner than the Livebox.
Secondly - there are a lot of diagnostics in a modern exchange. I'm surprised it took that long for the exchange (or an engineer) to probe the line and notice the mains present. If nothing else, I would have imagined that it would have presented a fairly distinctive hum...
...assuming that it didn't cause the phones of the subscriber who put the mains into the line to ring continuously. [back in my youth I made an effective phone ringing simulator by stepping the mains down to 110V (a UK to US (isolated) transformer) and then routing the output via a relay controlled by a BBC Micro; put that into the phone and it'll ring - used that for a school play]
I think if you look at the proposed rules in detail, what the FCC are concerned about are changes to the operation of the "radio" part of the device. The FCC wouldn't really care if hackers can change the rest of the software, provided that the software-defined characteristics of the transmitter can't be mucked about with. And I can see some merit in that.
The issue is that a lot of current hardware isn't really designed in such a way you can easily separate the ability to update the RF-related firmware from the ability to update the rest of the firmware - and it would probably add to the manufacturing cost if that capability were added - so I don't think it's the FCC's intention to shut down open source developers, but it may be an unintended consequence.
"The issue is that a lot of current hardware isn't really designed in such a way you can easily separate the ability to update the RF-related firmware from the ability to update the rest of the firmware - and it would probably add to the manufacturing cost if that capability were added ..."
The cost would be trivial. It just takes a few logic gates to carve out a block of address space for the radio processor, and another shared block for the user processor to drop updates in. The reckless vendors need to be stopped in their race to the bottom.
The real justification is that their router firmware is so crap it can often be hacked over the network. We are one malware away from having millions of wireless gadgets turn themselves into jamming bricks. This problem is a good justification for regulators to cram proper infosec isolation down the throats of device manufacturers. If some Chinese garbageware vendors go out of business, so be it.
Just it make also very simple for someone to modify the radio controlling parts. It would be just like the malware kits, if you had to implement it from scratch it would require a lot of expertise, if most of the hard work is already done, it takes very little to implement your bad app.
Really do you think that Nanny Britain or any other EU country is any less of a Nazi? Your handle is objectionable and should have been moderated.
However that link is for PDF documentation dated two years ago and comes from the EUC and covers harmonization of the radio frequency aspect specifically. You should already know how touchy governmental agencies are about radio interference especially since none of them can ever agree which frequencies are for public and which specific to governmental use only.
Far be it for ANY government entity for ANY country to have a clue about anything technical or how it affects anything. So they write rules and regulations in an idiotic manner and many times let manufacturers of equipment insert language that favors their viewpoints.
Methinks someone is so distrustful that not even wire mesh AND tinfoil is enough to keep the crazies from promoting distrust when the simpler explanation is just laziness and stupidity. I hear if you put your head in a microwave oven on high for 20 minutes, you won't hear the voices anymore.
I hear if you put your head in a microwave oven on high for 20 minutes, you won't hear the voices anymore.
Ah, except to do that you need to circumvent safeguards to operate the microwave with the door open. Or remove the head. Either of those requires unlicensed 3rd party modifications.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
