Asus locking down routers to comply with new FCC rules

[jegesq]

If your phone can't connect, my guess is that it's likely an issue with your phone, not the router.

Frankly, I don't really care what occurs in the EU or the UK. I'm in the U.S. and believe me, we have our own issues with the FCC.

My concern though, whether one is in the U.S. or EU, or anywhere else is that blaming a manufacturer because your particular environs or client devices may not be working optimally is silly if you understand the background, regulatory framework and technical requirements that the manufacturers are expected to meet in order to sell their devices within a region. As you say, if you're unhappy with your Asus router, then either RMA it, return it, or buy something else and see if that solves your problem. But really, if all you ever read are DJJHawk's two posts, then you'll understand that this is something that is applicable to any manufacturer, and it's a problem that is bigger than just what channels Asus decides to make available. So if you can find a SOHO router that makes more channels available without the need to change country codes, go for it and let us know how it performs. What do you have to lose?

 

[NGRhodes]

To be blunt their EU firmware is broken, and people using the a region is more widespread than you think.

If its broken, its not because its specifically EU.
No problems with the EU firmware on my AC56U, similar Wifi signal strength and transfer speeds as my old Cisco E4200 (running stock or tomato which I was running up until 3 weeks ago next to my AC56U), far from broken.

 

[Chrysalis]

I flashed tomatousb to see whats happening on that side, UK region removed, EU region added but unlike asuswrt the EU region allows higher channels, interestingly not channel 100 tho, had to use 104.

It is possible that my billion and zyxel are actually set to EU, I just assumed was UK as they allow the higher channels. It seems asus are the odd ones out here.

NGRhodes there is dozens of posts on this site about people having issues with the EU region, and they get fixed when they chose the a region. What reason also do you think the fork is so popular?

 

[RMerlin]

What I'd like to add to this endless discussion about regions is this: people who blindly switch their region to UK and call it a day should take a look at the router's source code sometime. The firmware doesn't have actual support for that region. A lot of the configuration that usually happen if the region is US or EU does not happen at all, as the firmware itself doesn't know about those regions. For instance, DFS only gets properly initialized if the region is set to EU.

In "fixing" one issue, they are creating quite a few others. Between all the "my wifi is unstable" posts this forum gets flooded with, I always wonder if those people are using a region that's actually supported by the firmware, or if they have hacked things to switch to some other region.

 

[RussellInCincinnati]

It sounds like people are complaining about 2 limitations, (1) max transmit power and (2) access to various 5 gigahertz channels.

===============
(1) Max Transmit Power Limitation Complaints

My experience is that 80 milliwatts is plenty of power for HOME ENTERTAINMENT and PERSONAL (NON-COMMERCIAL) distribution of internet access. That is to say it's pretty easy to distribute 120-250 megabits per second or so of internet bandwidth throughout a home with one or more inexpensive ($55 to $200 dollar) 80 milliwatt WiFi access points. Or with $70 dollar powerline (say Zyxel PLA5205) adapters that don't use wireless spectrum at all. There are 5 gigahertz repeaters (EdiMax BR-3208 433mhz) that are as little as $36 dollars.

The inexpensiveness of (powerline and repeater and ethernet cable) workarounds for an 80 milliwatt home router limit, the adequacy of 80-100 milliwatts, and the increasing availability of 3G/4G phone data networks, makes it feel silly to me to complain about a limit of 80-100 milliwatts for a router for home/non-professional use.

If you feel you "need" more than 120 megabits per second access to the Internet in your home (hmm I find 15 megabits per second to be pretty nice), because you are an unusual person or you have PROFESSIONAL/BUSINESS needs, nobody is keeping you from wiring your home or using powerline adapters or buying enough 5 gigahertz repeaters (160mhz bandwidth shortly becoming available) to get what you want.

Yes there is SOME cost to making people spend more money/lay wires/use powerline equipment to get more bandwidth at greater distances, because of regulatory power limits. But governments also correctly perceive there is some cost to allowing people to easily buy super cheap equipment...that has great amounts of transmit power that may be used in crowded urban environments in a way that interferes with other folks.

In short my opinion is that 80 milliwatts is "enough" inexpensively-available power for the MAJORITY of home/personal/entertainment networking and internet access. If you are a commercial user or one of a world minority of people who feels they "need" more than 120 megabits per second internet access across more than a few rooms of a home, then what is so terrible about making you heavy-networking folks buy powerline equipment or more repeaters or ethernet cabling? So that the great majority of us folks who are at less than 50 megabits/second and want to use dirt cheap WiFi iron, are not plagued by overly crowded spectrums...caused by too-widely-used greater-than-100-milliwatt WiFi users?

It is silly to make one-sided arguments that the government or Asus is limiting your freedom, without acknowledging there are some GOOD REASONS to limit the mass of folks' ability to easily flood high-density population areas with greater-than-100-milliwatt WiFi transmitters.

==================
(2) 5-gigahertz Channel Selection Limitations

If you really care about getting a clear channel in an urban area, it is not asking too much to ask people to use $36 dollar WiFi routers that can use 6 gigahertz channels. The range of 5 gigahertz is so limited that I doubt we will need many of those channels, even in crowded urban areas, to get a clear WiFi connection. We ESPECIALLY DON'T NEED TO WORRY ABOUT NEEDING LOTS OF 5GHZ CHANNELS IF WE DON'T LET PEOPLE BROADCAST ON THOSE SHORT RANGE CHANNELS WITH TOO MUCH POWER.

So who cares if we don't have bunches of 5 gigahertz channels available, as long as people don't have easy access to the kind of high-power transmitters that will cause any of us to interfere with our neighbors anyway?

==============
In summary even the most-limited-in-power-and-channel home WiFi equipment can give the vast majority of us adequate home networking at the cost of less than $40 dollars per WiFi transmitter. Heck TP-Link makes $28 dollar 2.4 gigahertz repeaters.

If we are part of the personal minority or business user set, that needs more range and bandwidth than a few 5 gigahertz channels and 80 milliwatts can provide, we are in a minority. And we should accept some extra cost/trouble/cabling/Powerlining in extending networking range and bandwidth, for the greater good of letting the majority of people in crowded areas have typical-personal-use cheap WiFi equipment and throughput.

 

[Chrysalis]

What I'd like to add to this endless discussion about regions is this: people who blindly switch their region to UK and call it a day should take a look at the router's source code sometime. The firmware doesn't have actual support for that region. A lot of the configuration that usually happen if the region is US or EU does not happen at all, as the firmware itself doesn't know about those regions. For instance, DFS only gets properly initialized if the region is set to EU.

In "fixing" one issue, they are creating quite a few others. Between all the "my wifi is unstable" posts this forum gets flooded with, I always wonder if those people are using a region that's actually supported by the firmware, or if they have hacked things to switch to some other region.

Merlin do you know why asus cap to 80mw, when 200mw is permitted?
500mw is permitted for the higher channels which they disabled.

 

[john9527]

Merlin do you know why asus cap to 80mw, when 200mw is permitted?
500mw is permitted for the higher channels which they disabled.

My understanding (limited as it may be )....the 200mW number is the total transmitted power, usually stated as Effective Isotropic Radiated Power (EIRP). If you convert the 200mW to dBm, it's about 23 dBm. Now the antennas have a gain, and contribute to this value. Depending on the reference, the Asus antennas are either 3.5 or 5 dBi gain. Let's use 4 for the calc. So the max power into the antenna is 23-4 = 19 dBm. 19 dBm converted to mW is 79mW.

BTW....if you do the same calculations for the #a mod...that raises the max allowable power to the antennas to 316mW. When you add the antennas and do the conversions (exercise for the reader), it results in an EIRP of 1W.

 

[jegesq]

Chrysalis:

Looking at the EU regulations, specifically the ETSI EN301893 V1.7.2 (2014-07) standards, they provide the limits for RF output power and power density at the highest power level (See Section 4.4.2.2).

For 5150 to 5350 (channels 36-64), the mean e.i.r.p. limit is 20 dBm (but 23 dBm if the nominal bandwidth falls completely within channels 36-48), without TPC being implemented. Also, it limits RF output to a mean e.i.r.p density limit [dBm/MHz] of 7 (or 10 if entirely within channels 36-48).

Since Asus is only exposing channels 36-48 in the EU, using John's math (e.g., 23dBm minus 4dBm for the average antenna gain), this is how you'd get to what Asus is using, i.e., ~80mW.

And actually, 200mW converts to 23.010299957dBm, so John was damn close.

So the real limit is not just "200mW" (which is the number bandied about on various blogs and white papers written by bloggers), but is actually the result of a calculation used to arrive at maximum RF output permitted.

 

[mavack]

Another thing that i'm not sure if anyone has taken into account about the wireless channels is that 802.11 devices are not the only devices to use 2.4ghz and 5ghz bands.

AV video senders, baby monitors, etc etc. An AP can't see these, the only way to check for interference is to use a spectrum analyser.

A total disregard for everything else within the band is what gets us into these problems, devices should dynamically adjust their power to be the minimum required to get the job done in the space.

 

[jegesq]

Another thing that i'm not sure if anyone has taken into account about the wireless channels is that 802.11 devices are not the only devices to use 2.4ghz and 5ghz bands.

AV video senders, baby monitors, etc etc. An AP can't see these, the only way to check for interference is to use a spectrum analyser.

A total disregard for everything else within the band is what gets us into these problems, devices should dynamically adjust their power to be the minimum required to get the job done in the space.

Don't forget microwave ovens, car alarms, USB 3.0 interference, Bluetooth interference (yes, BT uses the same 2.4ghz frequencies as wifi, just with much narrower 1mhz channels...and usually the interference runs towards BT non-functionality rather than vice-versa), home phone systems, etc.. They don't call this the "ISM" ("Industrial, Scientific and Radio") band for nothing.

I often wonder, as I read through the messages here about people complaining that their 2.4ghz or 5ghz radios are crapping out, or are behaving badly, how many of them have badly shielded external HD's connected via USB ports on their routers that are emitting too much RFI and causing the radios to malfunction. People wonder why Asus moved the USB 3.0 ports to the front on some of their newer routers, but I suspect it has a lot to do with interference generated by badly shielded USB devices. See, for example, this: http://www.intel.com/content/www/us...al-bus/usb3-frequency-interference-paper.html

 

[stevech]

"Wireless isn't a thousand times harder than wired, it's a million times harder!"

Professor Paulraj, Standford.

 

[Chrysalis]

My understanding (limited as it may be )....the 200mW number is the total transmitted power, usually stated as Effective Isotropic Radiated Power (EIRP). If you convert the 200mW to dBm, it's about 23 dBm. Now the antennas have a gain, and contribute to this value. Depending on the reference, the Asus antennas are either 3.5 or 5 dBi gain. Let's use 4 for the calc. So the max power into the antenna is 23-4 = 19 dBm. 19 dBm converted to mW is 79mW.

BTW....if you do the same calculations for the #a mod...that raises the max allowable power to the antennas to 316mW. When you add the antennas and do the conversions (exercise for the reader), it results in an EIRP of 1W.

yeah I know a is higher but on a I reduced in the GUI the power setting.

On tomatousb the signal is weaker than my merlin asuswrt region a setting but it doesnt break and higher channels still work.

I have sent some questions to asus support to ask what they playing at. It works like crap on their latest official firmware and its my most expensive router.

 

[jegesq]

yeah I know a is higher but on a I reduced in the GUI the power setting.

On tomatousb the signal is weaker than my merlin asuswrt region a setting but it doesnt break and higher channels still work.

I have sent some questions to asus support to ask what they playing at. It works like crap on their latest official firmware and its my most expensive router.

The point that Merlin made was that if you are setting your region to anything other than EU, you aren't implementing DFS or TPC, and thus are not operating on the other channels (where DFS and TPC are required) in a legal manner. And that's the case even if you lower your GUI's power settings. Lowering those settings does nothing to DFS, and indeed, DFS is outside the user's ability to control (other than by changing regions to one that is not authorized).

Perhaps Tomatousb is implementing DFS and TPC on those higher channels. I have no idea about that and you should address that to their devs (or perhaps Merlin or someone else knows).

 

[Chrysalis]

Some info here.

Currently my Samsung S5 will happily use the higher channels, there is no option to choose the region, but since I brought the phone in the UK, I am assuming its EU region.

Here is where it perhaps gets interesting.

I also recently acquired a oneplusone phone, this phone runs cyanogen a AOSP based ROM. In its default state it also happily used the higher channels. However I put on a 3rd party firmware and let it do a OTA update to a newer build of cyanogen and now it can only use the lower channels (basically doesnt detect my 5G wifi), its either the ROM update or the 3rd party kernel that made this change but until I test putting the stock kernel back I dont know which it is yet. But if it was due to the OTA update that is interesting as it was only an incrmental update not a major new version number of the ROM. On the phone I can adjust regions, and another thing of interest is that Germany has its own region but no other EU countries do, I tested all of them and none of them use the higher channels. So it looks like whatever the source of the change was, is they took the same approach as ASUS and said sod it, lets just can the higher channels.

Also ofcom did reply, they confirmed the UK is following EU guidelines.

 

[jegesq]

Phones that only operate in the upper band channels as "slave" devices, do not need to implement DFS or TPC so long as they are part of an RLAN that is controlled by a "master" device that is using DFS and TPC. If the phone only accesses the non-DFS/TPC channels (36-48), then DFS and TPC are of no concern. But to the point, since phones that access wireless RLAN's don't function as "master" devices (that's the wireless router's role), there's no technical reason why one phone's OS would provide access to the higher channels and another's OS not. (See ETSI EN 301 893 V1.7.2 (2014-07) 4.7.1.3.)

Thus, whatever reasons the independent third-party developers of your version of Cyanogen had, I doubt they were the same as the rationale of ASUS. Why whoever developed your third-party version of Cyanogen did what they did, who knows? You should probably e-mail the devs and see if they'll tell you why they only made the lower channels available.

As far as Germany having "its own region" that's not quite correct. What Germany did was originally implement even stricter requirements than those set forth in the earlier draft standards prior to ETSI EN 301 893 V1.7.2 (2014-07) (which individual member countries of the EU are permitted to do). Germany first required DFS and TPC to be used on the 5250 through 5350 GHz channels (channels 52- 64), and on 5470 –5725 GHz channels (100-140), and also limited 5150 –5350 GHz (Channels 36-48) to only indoor use. The same restrictions were later "harmonized" by the entire EU in 2005 and in 2007 (See, Official Journal of the European Union, COMMISSION DECISION of 11 July 2005, and Official Journal of the European Union
COMMISSION DECISION of 12 February 2007 amending Decision 2005/513/EC, and they are now embodied in the full EN 301 893 V1.7.2 (2014-07) document.

Simply put, Germany is equally bound as a member of the EU by the ETSI R&TTE Directive and it doesn't have "its own region" any more than the UK has its own "region".

 

[jegesq]

BTW, I haven't really researched this thoroughly, but I believe the ETSI requirements also provide that if a "slave" device (such as a phone) is also capable of functioning as a "master" (that is if the phone is capable of operating as a wireless "hotspot", and it is using channels above 48, i.e., those that require DFS and TPC that the phone is also required to have the ability to implement DFS and TPC on its own. To the extent third party firmware allow access to the upper band channels and doesn't also implement DFS and TPC when the phone is operating as a wireless hotspot, that would not be legal in the EU.

Edit to add: Yes, it's right there in ETSI 4.7.1.3 as well:

"A device which is capable of operating as either a master or a slave shall comply with the requirements
applicable to the mode in which it operates.

"Some RLAN devices are capable of communicating in ad-hoc manner without being attached to a network. RLAN
devices operating in this manner on channels whose nominal bandwidth falls partly or completely within the frequency
ranges 5 250 MHz to 5 350 MHz or 5 470 MHz to 5 725 MHz shall employ DFS and shall be tested against the
requirements applicable to a master."

 

[djjhawk]

Don't be one of the boneheads

It appears I may have been pointing the finger in the wrong direction in my original assessment. Nonetheless, I certainly underestimated the magnitude of the problem in the EU region.

http://ec.europa.eu/enterprise/sectors/rtte/files/5th-rttems-report_en.pdf

http://www.cept.org/files/1051/Tools%20and%20Services/Public%20Consultations/2013/Draft%20ECC%20Rep192%20on%205GHz%20DFS.docx

 

[Chrysalis]

jegesq I am referring to EU having its own setting on the phone.

Anyway my router now works on that phone again, I am guessing was a temporary issue.

Also I got a 2nd reply from ofcom today, they state we are currently 100% in line with EU but its currently under review as they are considering relaxing the rules and I have been invited to give them feedback.

Also I have queried ofcom if its legal for OEM's to remove channels as the document on their site mandates those channels need to be available.

 

[pirx73]

And with regard to Nigel's comment about his purchase of a Nexus 7, all he was pointing out was that his Nexus 7 only had access to channels 36-64, and the channels from 100-140 weren't supported either.

Channels from 100 and above and mostly used for outdoors hence allowed 1W max transmitting power in EU.
Nexus 7 being "indoor device" technically (meaning not designed for constant outdoor usage in all possible weather conditions) does not support those channels.

 

[jegesq]

Also I got a 2nd reply from ofcom today, they state we are currently 100% in line with EU but its currently under review as they are considering relaxing the rules and I have been invited to give them feedback.

Everyone (all of the EU as well as the FCC) is looking at determining whether DFS and TPC need to be implemented on all channels in all countries, and all of them are also looking at opening up more channels as well. But for now, it is what it is.

Also I have queried ofcom if its legal for OEM's to remove channels as the document on their site mandates those channels need to be available.

Of course it's "legal" for OEM's to not make certain channels available on devices they sell. The fact that certain frequencies exist and the EU's ETSI regs have stated that they "may" be used (i.e., that they are available for use) doesn't mean that they "must" be used by manufacturers. But if they are used, that usage has to comply with the R&TTE Directive and the rules promulgated by the EU.

But I'll be interested to hear what you learn from Ofcom. Please let us know.