Linksys WRT Routers Won't Block Open Source Firmware, Despite FCC Rules (arstechnica.com)

The FCC requires all manufacturers to prevent users from having any direct ability to change RF parameters (frequency limits, output power, country codes, etc). The easiest way for a router manufacturer to comply with FCC's guideline is to block the open source router firmware -- which is what TP-Link has been doing. But thankfully, at least one router manufacturer doesn't think blocking the firmware is the right way to go about it. Ars Technica reports: Linksys has been collaborating with chipmaker Marvell and the makers of OpenWrt to make sure its latest WRT routers can comply with the new rules without blocking open source firmware, company officials told Ars. Linksys' effort stands in contrast with TP-Link, which said it would entirely prevent loading of open source firmware on its routers to satisfy the new Federal Communications Commission requirements. "They're named WRT... it's almost our responsibility to the open source community," Linksys router product manager Vince La Duca told Ars. Cybersecurity experts have urged the router manufacturers to not block open source firmware.

Comments

Excellent

[Anonymous Coward]

Excellent decision. I know what my next router will be.

Re:Excellent

[negRo_slim]

While I'm not sure I'll go with Linksys, I am certainly disappointed I picked up a TP Link right before all this came to light. At least I can enjoy knowing that most people only buy TP Link to flash custom firmware and their foolish decision will most likely hurt their bottom line.

I still don't understand what lead to all this, I highly doubt putting a few more milliwatts out or using a few MHz on either side of the allotted spectrum was causing much (if any) actual harm.

Re:Excellent

[Anonymous Coward]

Don't be deceived, it's still possible to flash custom firmware on a TP-Link device even after all their work. You can find more information via search on the openwrt forum.

Re:Excellent

[Anonymous Coward]

No. Fuck no. People are morons and we must not let people decide if it's OK to "[put] a few more milliwatts out" or using parts of the spectrum they're not supposed to. That's begging for localized shouting matches and overuse of allocated spectrum... where both are problems we already have.

The FCC absolutely did the right thing here, and so is Linksys.

Re:Excellent

[Solandri]

I still don't understand what lead to all this, I highly doubt putting a few more milliwatts out or using a few MHz on either side of the allotted spectrum was causing much (if any) actual harm.

The "open" 5 GHz band spans from 5030 - 5835 MHz. Right smack dab in the middle of it at 5600 MHz is terminal doppler weather radar. That's one of the radars used to make the pretty rainfall pics you see on the news reports. More crucially, airports use it to detect wind shear conditions which have brought down several airliners in the past.

The problem is, because it's doppler radar, the frequency shift of the radar signal is crucially important, not just the return signal strength. So the FCC has carved out a band from 5250 - 5730 MHz and declared it DFS - dynamic frequency selection. Equipment can use these open frequencies, but if they do they have to monitor to see if weather radar is being used, and immediately shift to a different frequency outside this band if they detect weather radar.

A few routers do implement this frequency shifting. The vast majority simply block out those frequencies in their firmware (which is why your 5 GHz channel selection is limited to channels 36-48 and 149-165). But most third party firmwares are made in countries which don't have airports with TDWR radar (in fact most U.S. airports don't yet), so their governing agencies don't restrict these frequencies, so the firmware authors make no effort to limit use of these frequencies. Loading the third party firmware onto a router in the U.S.allows your router to spam these frequencies indiscriminately, leading to TDWR possibly being unable to detect wind shear or detecting it later than it could have without the interference, possibly causing another airliner to crash.

Re:Excellent

[Anonymous Coward]

Excellent explanation. Visually, devices that interfere with TDWR displays appear as a wedge that is several degrees wide. The wedge obscures everything from the center of the display to the edge. Here's an example:

http://fpvlab.com/forums/attachment.php?attachmentid=69438&d=1452686140

Re:Excellent

[EndlessNameless]

There is finite usable spectrum, and Wifi had to go somewhere.

Usable spectrum in this case means:

* not absorbed by the atmosphere
* reasonable effective range per unit power
* harmless to organic tissues
* capable of penetrating typical construction materials

If you're aware of an unallocated band that you think the FCC has overlooked, feel free to contact them to ensure the oversight is corrected.

Re:Excellent

[danbob999]

The problem with the new WRT routers is that they sell for twice the value of comparable hardware from other vendors. And if you look carefully, you can find models which are just as good, if not better, in running 3rd party open source firmwares such as OpenWRT.
The original WRT54G was cheap, that's why it succeeded.

Re:Excellent

[The-Ixian]

Yet you still are supporting their practices by buying a TP-Link device....

Re:Excellent

[epyT-R]

Radio is not reliable when shared, regulation or not. It only works when the players work together perfectly. When does that ever happen? It's like cooperative multitasking os's expecting applications to behave. It's a fool's game. If this radar is crucial, maybe we shouldn't share this band with consumer electronics. The FCC failed hard here.

Re:Excellent

[Solandri]

* not absorbed by the atmosphere

Actually, 2.4 GHz and 5 GHz were chosen because they're absorbed more by the atmosphere. 2.4 GHz matches up closely with the resonance of water molecules (which is why microwave ovens operate at 2.45 GHz - they heat up food by pumping energy into the water molecules). And 5.6 GHz is absorbed readily by raindrops for some reason I haven't yet been able to learn (which is why it's handy for weather radar).

Broadcasts on both bands thus attenuate more quickly relative to other frequencies. This not only makes them less useful for long-distance transmissions, but it makes them ideal for unregulated transmissions. The signal blasted by an unregulated device at these frequencies doesn't travel as far through the atmosphere, so the radius of its noise footprint is smaller compared to a different frequency. Basically, you can squeeze more devices closer together without interfering with each other at these frequencies. Both of these "features" make them ideal choices for open frequencies.

Re:Excellent

[mattventura]

The reason it's an issue is because in the typical race to the bottom, a combination of things is happening:
1. Routers are more often being built using an SoC that includes (at least one) wireless chipset.
2. Wifi cards are more often being built without builtin firmware storage, meaning it has to be loaded at runtime.
Thus, you've got a device with a wireless chipset whose firmware is either stored along with the main firmware stoarge, or has to be loaded by the OS. Think of it like a computer where the ethernet firmware was stored on the hard drive along with the OS. It would be difficult to make it impossible to modify the ethernet firmware without also preventing the user from installing an OS of their choice.

Compromise

[Espectr0]

Firmwares that ask the country you are in. Choosing US greys out the appropiate sections. If clients lie, then the company shouln't be at fault.

Why was there ever really any doubt...

[mark-t]

... that some router manufacturers were going to do this?

Re:A good decision

[Anonymous Coward]

This action is not a violation of the FCC rule in question. That rule defined a chain of responsibility for purposes of assigning blame when and only when someone is using their wireless router beyond the scope of FCC approved frequency/intensity ranges.

For a fine to come about, someone has to flash the firmware (easy it seems) with another firmware that does not respect the FCC mandated boundaries (I think that would require changing the source and recompiling for most of them) and then pick a frequency and power level that is prohibited for WiFi use.
While I forget the order of blame that the FCC decided on, this announcement indicates that Linksys has run the numbers and determined that the increased fanaticism from OSS enthusiasts for their routers will more than offset any anticipated culpability when people misuse their hardware.

Summary

[DrYak]

The FCC rules do not mandate, either directly or indirectly, that OS firmware be blocked.

Which was already explained in the Summary.

I know this is /. and nobody reads TFA, but you could at least read a little bit past the title and at least read the summary.

A title has limited number of words and characters: "despide FCC rulins" is as mush as the author managed to cram into it.

And the FCC *IS* causing most manufacturer to lock their firmware. FCC doesnt madante litteraly against opensource and user-upgradeable firmware, but the ruling is strict. No unlawful signal shall be emitted by a device. Under no circumstances.

- Either a manufacturer has to jump through some complex hoops to find a solution which both open and user accessible (following the spirit of the GPL license used by some component. And litteraly following GPLv3) AND at the same time prevent the end-user from emitting signals that fiolated FCC rules (e.g.: emitting on a EU-only frequency in USA).

- Or the manufacturer could just lock everything in a box, and only let cryptographically-signed firmwares in, and call it a day. (and hope no GPLv3 got violated in the process).

Nearly every constructor goes for the latter. Only LinkSys Fritz and a few other go for the former

Re:"Despite FCC Rules"

[amicusNYCL]

It literally says in the first sentence of the summary that the FCC rule requires that manufacturers prevent users from changing the RF parameters. There's no claim that the FCC rule says that you have to block OS firmware. It should be obvious to anyone that the easiest way to prevent users from changing the RF parameters is to block OS firmware (that's the second sentence of the summary, by the way), but the story is that Linksys is going to find another way to lock down the RF module without blocking OS firmware.

Features

[Frederic54]

I do not not install a 3rd party firmware to use channel 13 or crank power output, but for features:

• Wireless bridge
• OpenVPN client
• OpenVPN server
• Powerful QOS
• IP traffic monitoring
• IPv6
• Multiple SSID
• Full USB support (printer, HDD, 4G, etc)
• VLAN
• MultiWAN
• etc

All this on a $79 router.

Re:Features

[wierd_w]

Some devices are even cheaper. The ZSUN sdcard sharing adapter is only 11$ off amazon. It can run openwrt, meaning basically all of your listed features are possible (vlan not so much, only 1, unconnected, physical ethernet header).

I has 2 wifi radios, so it makes a very cheap repeater, and has a reasonably fast internal SDCard slot, so it could do light duty fileserving. With a little hardware modification, it could do USB print serving as well. (Designed to be powered from a USB port, but is wired in host configuration, not gadget. This means that if you supply power to the voltage pins, and put a gender changer on the connector, it should be able to see a connected printer and still operate.)

Mostly the FCC just doesnt want to deal with people in the US telling the router that they live in Japan, and thus using spectrum on the far edge of the 2.4ghz and 5ghz bands that are not public license in the US.

The chip makers are cheap fucks and control the radios with a volatile software blob (looking at you broadcomm) which means that an open platform like linux is able to push custom blobs on at boot time. Because the chip makers wont right thier wicked ways, the router oems just lock the device down in order to comply. The FCC would have done better to go after broadcomm and pals, if you ask me. Marvel, mentioned in the article, does things right with actual honest to god firmware controlling the radio.

Re:As a European, I am worried.

[wolrahnaes]

We get to use more frequencies than the US. I've owned wireless devices which couldn't connect to networks on channel 13, because they were hard locked to US rules, despite being sold in Europe. I sincerely hope that they only strictly enforce a chosen set of rules, instead of enforcing which set of rules the users can choose.

Don't use channel 13.

Look here: https://en.wikipedia.org/wiki/...

1, 6, 11, and 14 where available are the only channels anyone should use. They allow the most non-overlapping use of the available spectrum. If you use anything else, you've now limited the choices of everyone around who wants to avoid interference.

IMO AP vendors should really lock down the channel selection. The legitimate reasons to use any other channels are so rare and specialized that it at least shouldn't be easy to do.