Re: [Openbts-discuss] Device doesn't see network at all (openbts-uhd + usrp2 + wbx)

[David A. B. <dbu...@jc...>]

Alexander -


On Jun 13, 2011, at 3:28 AM, Alexander Chemeris wrote:

> Hi David,
> 
> As a long term supporter of OpenBTS community and the open-source
> telecom in general, I can't help but object. Aggressive approach like
> that doesn't work well for a community.

And as you are a long term supporter and actual contributor, I respect your comments.

> 
> On Mon, Jun 13, 2011 at 06:54, David A. Burgess <dbu...@jc...> wrote:
>> 
>> OK, so here's a message to everyone who is trying to run OpenBTS on a stock
>> USRP and trying to get their everyday cellphone to camp to it: STOP BEING STUPID.
>>  What Kurtis suggests below is a strongly recommended mode of operation: Do not run
>> OpenBTS in a band that is also used by public networks in your area.  I say this routinely.
>>  I put it in a wiki.  Nobody is listening. At least once a week, some kid trying to run
>> OpenBTS on a stock USRP in the same band as their local cellular carriers posts to
>> the list saying "OpenBTS doesn't work". It doesn't work because you didn't bother to read
>> the wiki or understand the technology before turing on a radio transmitter in a licensed
>> band, a band used to provide a critical public service, and, whether you know it or not,
>> inviting all of your neighbors' phones (not just your own) to jump on to your little network.
>> (See my other recent post, "you were warned".)
> 
> You will be surprised how many people actually read the wiki
> carefully. I know, because they point to it when they buy our
> ClockTamers. The amount of questions we're still getting means only
> that community is growing really fast. Solution for this is to create
> yet another wiki page or an entry to FAQ which say this clear and
> concise message, so instead of typing the answer again and again, we
> could just answer with a link.
> 
> Another great thing for the community would be a list of all possible
> problems with a first-time setup of OpenBTS. There are much more of
> them then just a clock issue - e.g. Asterisk setup, antennas, etc. A
> concise and to-point troubleshooting guide for newbies will solve a
> good deal of questions on this list. I'd love to see someone taking
> the task to create at least an initial version of such document.

I think this is a great idea.  The wiki is publicly writeable.  I'd love to see someone taking that task on as well.

> 
>> Besides making clocking much less of an issue, operation in a non-standard band also
>> greatly decreases the likelihood that you will end up in prison or paying heavy fines for
>> disrupting a public network.  This is no joke and your local public cellular network is not
>> a toy.
> 
> I should note that, operation in a non-standard band for your country
> may impose much more threat on you. They're non-standard for a good
> reason - something else is operating there. Probably a military
> system. Probably a public safety system. Probably an aircraft
> communication channel. By operating in a non-standard band without
> researching the topic for your exact location you can break something
> much more important then just a public cellular network. Consequences
> my be much worse if you break, e.g. some military wireless link.

All true, and it is very important to know your spectrum before operating.  However, the services you list there usually operate at power levels well above those seen from a USRP and unless you live next to a military base or airport the receivers will be far away.  Of course, that doesn't make it legal or OK, but simple, ordinary radio interference is a separate issue from the kind of "false basestation" risks that are unique to cellular experimentation.

> 
> Then, you don't take into account differences between countries. E.g.
> in some European countries you can legally transmit in some part of
> 1800 (IIRC) band, and in other countries regulations may be different.
> So it looks like you're trying to enforce something based only on US
> law and regulations. This is a wrong way for an international
> community.

I'm not saying something based on US laws and regulations.  I'm saying that inserting yourself into the network side of a public network without understanding the consequences is unethical, regardless of what the laws are in your area.

> 
>> This kind of clueless, careless  operation is what makes me constantly question
>> continued support for the USRP in the public release: it makes the barrier to access
>> too low to keep out irresponsible people.  When someone runs OpenBTS in their local
>> public-carrier cellular band (usually just because they are too cheap/lazy to get a proper
>> multi-band unlocked phone)
> 
> Her is an interesting technical question. Most if not all of my test
> phones are quad-band. What is an order of bands scanning for a phone
> when I turn it on? If it starts with a band which it used previously,
> then your suggestion is useless - phone will find your incumbent telco
> signal first and will not able to see your signal. To really get your
> suggestion working you probably need a phone where you can choose a
> band of operation. And this is much harder then just a "multi-band
> unlocked phone".
> (also my comment about disrupting non-cellular networks apply)

So get a phone where you can manually control band selection.  Many do allow that.  Some are cheap.  For example, the Nokia 8890 in field test mode allows you to manually disable either the 900 or 1900 band, leaving only the other band active.  Or the Treo 650, which allows you to manually select a 900/1800 or 850/1900 bandplan.  These are old phones and can be found used for reasonable prices.

> 
>> they are creating a real threat to public safety and those of us responsible for offering
>> this source code to the public do not care to be a party to that.  I have tried before
>> to give friendly advice to push people in the right direction and already know from
>> experience that it is a waste of time, so the next someone disregards this advice I will
>> try something new: deleting threads from the archive and banning people from the list.
>> 
>> To the rest of the OpenBTS community, I apologize, but we need to start enforcing
>> some standards for responsible operation.
> 
> Enforcing is a bad word for the community. Keep in mind you're not a
> law enforcement agency and you run an international community.

I'm not trying to enforce the law, but I am obligated to speak against unethical operation and maybe even take action.  Creating a threat to public safety is unethical, regardless of your nationality.

> Stop crying, stop enforcing - create a traction towards the right
> direction.

Crying?  It is very easy to talk like that when your head is not the one that will be on the block.

> People like Harald could help in understanding what is the
> right thing to do to create this traction. Probably more good
> documentation would help.

More good documentation would help, yes.  The wiki is publicly writable.

> 
> Especially bad would be if you start banning newbies and thus create a
> hostile environment for newcomers.

I get the impression that a lot of newbies are out there playing with fire and don't realize it.  This problem is growing faster as the project grows, since most of the early users were already more knowledgable about GSM than many of the people coming in now.

> I strongly recommend you to rethink your community relationship model
> and do more facilitation, steering and goal-setting work rather then
> punishment. It would be much more useful for the community if we see
> more mails from you with the former mood rather then with the latter.

I'd like to have a better "community relationship model" and will try not to react quite so harshly in the future, but am still very concerned about the potential for irresponsible behavior.  My harsh scoldings are probably bad for the project, but an incident where someone is harmed because of reckless behavior would be much, much worse for the project.  We need, as a community, a way to avoid that.  If you don't like the word "enforcement", pick another, but the "community" needs to get serious about educating these new people.  And "community" has to mean more than just you, me and a small handful of other people.

> 
> -- 
> Regards,
> Alexander Chemeris.


Best regards,

David