Low cost open GSM base station for developing countries
What did you do during your summer holidays? I was
building my own GSM network!
I have been waiting almost 20 years to make this phone
call!. Back in the 90's when I started my engineering
studies and discovered GNU/Linux, dreamed with a system
that could do what I have right now on my table :). The
first attempt was in 2001 when I was completing my PhD at
KTH in Sweden. During those years a group of researchers
in the Laboratory started to explore the possibility of
replacing as much of the GSM architecture (BSC, MSC) for
open source software. Although we had a commercial GSM
container with plenty of noisy equipment, the project got
stuck trying to find documentation for the GSM internal
interfaces.
In December 2008, I attended a conference about Mobile
Service in Developing Regions. I was very surprised not
to see anyone addressing the need of opening the GSM
infrastructure (hardware and software). Christmas period
was a good time for new projects so I decided to get in
touch with the openbts project and try to build their
base station. Assembling all the necessary hardware took
a bit of time! David Burguess from openbts has been
really helpful with hardware purchases!.
Yes! It does work, and that first phone call felt like a
different one :D. The base station uses a piece of
hardware known as USRP. The USRP is a radio communication
system where components that have typically been
implemented in hardware are now implemented using
software in a PC.
We are looking for sponsors for a deployment in a
developing region! Interested? Drop us a line!
-aep
PD. Thanks to the openbts folks for a great project!
Dear Alberto,
I'm very interest for your GSM base station for developing countries. I'll like to know more about and look how I can do it in Cameroon (Central African Region).
Waiting your prompt reply. Thanks. Jonas KEMAJOU SYAPZE Directeur OPED - Organisation pour l'Environnement et le Développement Durable B.P. 12 675 Yaoundé - Cameroun Tel: +237 77 60 23 83 / 22 00 35 06 E-mail: oped...@yahoo.fr / jksy...@yahoo.com --- En date de : Lun 24.8.09, Alberto Escudero-Pascual (lists) <aep....@it46.se> a écrit : |
It may well turn out that we will now start seeing some small change
in this regard. :-D
100% awesome Alberto!
How hard would it be to bridge that GSM stream with a VOIP network?
I ask out of interest in the two hybrid models:
-> Localized GSM network and using VOIP for backhaul to service
providers.
-> Localized VOIP network and using GSM for backhaul to service
providers.
*wicked.grin*
Someone should probably tell the regulator that the headaches w/ WiFi
are nothing on what will be coming down the pike!
- antoine
--
http://7degrees.co.za
"Libré software for human education."
They set up this openBTS on the "burning man" festical a year or two
back. Then they connected an Asterisk with VoIP to it and to the leased
lines. They had it up and running and suddenly noticed that people were
using it to call home.... (the GSM phone just attaced to the network, no
more questions asked about roaming....)
> -> Localized VOIP network and using GSM for backhaul to service
> providers.
>
>
This is is very easy to do today and no need of the openBTS. You get
TeleS GSM gateways (4 channel for about 750 euro) or other cheaper
models They do direct SIP Trunking and even analog ports if you want to.
Then just point your OpenSER og Asterisk to a trunk to the GSM gateway
and you are up and running. There are larger communitites connectiong a
local PBX with hunderds of phones to a GSM "backbone".
Sjur
VoIP expert looking for new projects!
http://www.usken.no
I personally believe that GSM is the best candidate technology low-
cost rural service, given an implementation of the network that is
properly adapted to the application. The problems with it are purely
regulatory, not technical. The core question is whether or not
governments are serious about universal service. Telcos have long
had economic excuses for limiting service and governments have
accepted those arguments. It will be interesting to see what happens
when those excuses are removed, and politics and greed will have to
be used more nakedly to maintain the current situation. I keep a blog
and I put all of my standard arguments and speeches there now:
http://openbts.blogspot.com/search/label/universal%20service
-- David
On Aug 24, 2009, at 3:37 PM, David Rowe wrote:
>
> GSM is a fine system: the handsets are cheap, and unlike Wifi it
> propagates well over long distances without line of site issues.
> Now we
> can make even GSM base stations using open source hardware and
> software
> components. The only problem is that GSM seems to attract business
> models that are too expensive for the people who really need them.
>
David A. Burgess
Kestrel Signal Processing, Inc.
-- David
On Aug 25, 2009, at 3:21 AM, Antoine van Gelder wrote:
>
> -> Localized GSM network and using VOIP for backhaul to service
> providers.
OpenBTS does exactly thi already. Direct GSM->VoIP was the whole
point of the project.
>
> -> Localized VOIP network and using GSM for backhaul to service
> providers.
That's a different problem. Probably not hard with Asterisk and a
GSM modem.
>
-- David
Femtocells: Vendor adopts 'direct to consumer' sales strategy
By Paul Rasmussen Comment | Forward
In an unusual switch, the Edinburgh-based femtocell developer, HSL,
has adopted a sales strategy that involves cutting out the mobile
operator and selling the micro base station direct to the consumer.
This move, seen by some as high-risk, comes after Vodafone's launch of
its own femtocell product was received by subscribers with huge
disinterest. Also, given the marketing power of Vodafone, HSL is
strangely not attempting to undercut on price--its product, with a
rather consumer-friendly branding of HSL 2.75G, is priced very similar
to Vodafone's Home Gateway device at around €180.
However, the company claims that its femtocell can work with any
network operator, or even multiple network operators, over a user's
own broadband connection. But, to do this, HSL needs radio spectrum in
which to operate, and integration into the network-operator's back end
infrastructure--neither of which is trivial.
HSL believes it can overcome these hurdles by persuading possible
consumers to register their interest on the company's web site. Using
this, HSL plans to convince mobile operators to route calls through
their network--why they should provide this service is not detailed by
HSL.
However, MD and technical director of HSL, Mark Hay, said that the
installation and use of the network would not cost operators very much
and would allow them to manage their service better so that customers
receive the best coverage that is possible given their locations.
The company has confirmed that it is pursuing agreements with
operators in Europe, the Middle East, Asia, Africa and Oceania with
regard to its femtocell product.
I think they can emit less power at least, i.e. be cheaper, as amplifiers
and other RF stuff is actually much more costly then DSP part.
> The USRP board provides nice scanning features that allows you to choose a
> channel that is not occupied. Picture attached for the curious!
Yeah, I'm using this to choose ARFCN for me too :)
But I think we should rather use beacon scanner - something similar
to what mobile does when searching for BS. Then we'll be able to get
a full list of BSs and used ARFCNs nearby and choose a free one
(semi-)autoamtically.
Also one should keep in mind that sequential ARFCNs can't be used
becauase of frequency band overlap - ARFCNs are spaced by 200kHz,
but have bandwidth of 270.833kHz. I.e. if you see ARFCN 5 occupied,
you cannot use ARFCNs 4 and 6.
And, sure, you should keep in mind, that license holders still will be
very unhappy and you may get a big troubles with autorities if they
find you.
--
Regards,
Alexander Chemeris.