GPRS data question

I'm trying to get some information on a proposed GPRS data link but
though the application is simple in principle I'm really struggling to
get reliable answers to some questions from Orange, Vodafone etc.
Here's the scenario:

1. I have a scientific instrument at a remote site which I'd like to
communicate with across GPRS from a central office.

2. There won't be a PC at the remote end, instead the plan would be to
interface the instrument via RS232 to a GSM/GPRS modem. I can imagine
that this would need a SIM card with a fixed IP address, assuming such
a thing exists.

3. I want to communicate from the central office PC via the Internet,
so I'm imagining that my IP packets would zoom off on to the net,
encounter a gateway somewhere - presumably run by the company
provisioning the GPRS data link - and thence on to the GSM/GPRS
network to arrive at the remote modem.

This all sounds like it should be do-able to me, indeed do-able
without too much special knowledge or configuration. Can anyone
confirm that I can do this and/or point me towards where I might try
to get some specialist advice - my email address is valid if anyone
wants to approach me directly. I was wondering if there's some
specialist contact numbers for the main mobile telcos for this sort of
application that would bypass the flummoxed 'agents' on the general
contact numbers. (I would say that I'm looking for a standard and
low-cost solution here, but the bandwidth required is also pretty
low.)

---

› See More: GPRS data question

On 08 Jun 2007 22:23:55 GMT, Gordon Henderson
<[email protected]> wrote:...

Many thanks for taking the time to reply but I what I described is
really what I want to do.

The GSM approach would certainly work OK and indeed
I've had that sort of configuration working in the past. But the
problem is that every time you want to request/send new data then a
new call must be set up. And if I want to pass new data every 10 secs
say (it might be more often) then the costs become prohibitive.

This is a scenario where just a few bytes of data need to be sent very
often. The total bandwidth required isn't huge, it's the frequency of
transmission that is the key requirement. AIUI (which may be faulty)
GPRS ought to be a good solution to this sort of requirement.

But overall I'm looking for a way to tunnel an RS232 link across GPRS
to set up a transparent data link. There's no way of reprogramming or
repackaging the way the data link operates I'm afraid. And there can't
easily be any GSM/GPRS device at the central end hence the requirement
that the central end sends/receives across the net.

This is the sort of data link that is very easy to set up nowadays
across a standard LAN or WAN - I'm struggling to believe that there's
not a simple way to use GPRS as part of this sort of link too.

JGD

On 08 Jun 2007 22:23:55 GMT, Gordon Henderson
<[email protected]> wrote:

>I've used these:
>
> http://www.westlake.co.uk/Cellroute_...al_gateway.htm
>
>more for telephony than data, but they support GSM data, and sending
>text messages. (and the have a built-in batery too, and are cheaper than
>Siemens M35 GSM data terminals which I've used for SMS-only purposes)
>
Useful link, Gordon, it does look an interesting unit. The PDF for it
mentions that it will take an external antenna with an SMA connector.

Thinking about installation on an island, less than 15 miles from a
transmitter site serving both 900 and 1800 operators. In-building
penetration is pretty hopeless but quite good signal outdoors. I was
wondering what sort of length limit would apply to the antenna cable,
and what sort of size would a suitable antenna be. Do you happen to
know if a 15 metre cable through an attic to an external antenna might
be too long?


______________
regards,
Ron

On 09 Jun 2007 13:38:44 GMT, Gordon Henderson
<[email protected]> wrote:

> This is the
>application these units are actually designed for - remote GSM/GPRS mobile
>phone coverage. I have one connected to my office PBX and connected via
>rs232 to my monitoring PC for sending TXT messages - I can also make
>outgoing calls via it, or accept incoming calls.
That sounds good. We use the Orange World Internet GBP 1 daily
service as a backup for internet access. When you say you make o/g
calls, did you mean data calls as well? Presumably so, if you can use
the data port for SMSs. And if you can send SMS, then sending 'World
1' to Orange (which triggers the Internet service for a day) would
work, although you can also place a voice call on 450 to Orange to
trigger the service as well, that would work via this setup if you
connected an analogue DTMF phone. Once set up, and connection
established using the AT+CDGCONT=stringthingy, the normal GPRS IP
service is presented on the data port, is it?

>It also has internal rechargable batterys - so in the event of a power
>failure you can still have a bog-standard analogue phone plugged into
>it and it will still work.

This is seriously useful in these locations. Although reticulated
power is available, power drops even more frequently than the normal
phone service does (phone off maybe 6 times a year, power off maybe 12
times a year). Sometimes, though, the GSM BTS also goes off less than
a couple of hours after a power cut, so if there's a wide area outage,
then a backup service isn't available for long.

Incidentally, sorry to go off topic for a minute, I don't think that
those users who ditch their fixed line provision always realise that
mobile providers have neither a universal obligation, nor any service
continuity obligation. This means that mobile service could be
interrupted for fairly long periods if, for example, the maintainers
cannot agree whose equipment failure has led to a loss of backhaul.
(GSM service was off here for 8 days on one occasion while different
engineers each 'proved' their section.)


______________
regards,
Ron

John Dann wrote:
> I'm trying to get some information on a proposed GPRS data link but
[snip]

Have a look at <http://www.itegno.com/product_front.html>.

If you need a developer / bod to put an end-to-end solution together
then I can put you in touch with someone. Depends where you are.

Regards,

--
Bruce Horrocks
Surrey
England
(bruce at scorecrow dot com)

On 11 Jun 2007 08:05:08 GMT, Gordon Henderson
<[email protected]> wrote:

>These are still just modems - so the OP still needs that magic black box
>to take rs232 from his data equipment and interface it to a 2nd rs232
>link running PPP to talk via these modems...
>
A bit like one of those old-fashioned X25 PADs, except talking PPP
instead of X25. Actually, a bit more complicated than that, because
wouldn't this magic black box also need to do connection supervision
(connect, disconnect, re-connect, re-attempt)?

Presumably the 'standard' asynchronous terminal servers expect a fully
functioning IP network behind them. So maybe a simple laptop running
linux could emulate that - present an IP interface on a network port,
and use a GPRS connection using some kind of PPP dialler? Plenty of
router distributions out there to do that - Freesco, for example, if
the GSM terminal has an RS232 port, or even Knoppix if the phone has a
USB connection. Still a 3, no, 4 box solution though - the OP's data
generating box, a terminal server, and a laptop - plus the GSM/GPRS
terminal.


______________
regards,
Ron

I'm pretty sure that there are (small) single box solutions (maybe
many) already available, ie literally a GPRS modem with an external
RS232 port and presumably a TCP/IP stack within the 'modem'.

The problem I've got, which prompted the original post, is that they
seem to be part of customised high-cost solutions. But if they work
for high-end solutions then there's perhaps no reason in principle why
they shouldn't work for lower-end - ie lower-cost - solutions - the
hardware cost isn't really the issue but it's the how-to knowledge and
the provisioning of the GPRS service which seems to be costing.

My take on this at present is that the companies that specialise in
GPRS data links are making too much money from their existing
customers to readily offer cheaper options. This isn't really a dig at
them - a good business will make money wherever it can and nothing
wrong with that. But at some point competition should kick in and
progressively bring in lower cost solutions. Unfortunately I can't
find any evidence of this happening yet with GPRS data links. The post
was primarily to see whether I had overlooked some relatively
inexpensive and readily available options, but the answer thus far
seems to be no.

JGD

On Mon, 11 Jun 2007 13:39:27 +0100, John Dann <[email protected]>
wrote:

>I'm pretty sure that there are (small) single box solutions (maybe
>many) already available, ie literally a GPRS modem with an external
>RS232 port and presumably a TCP/IP stack within the 'modem'.
>
>The problem I've got, which prompted the original post, is that they
>seem to be part of customised high-cost solutions.

Are these the sort of things you've seen already? I'm not completely
sure whether they fit your need, but they look to have the basic
functions.

http://www.bauschdatacom.be/viewprod...&catalogusID=1

http://www.harvest.com/modem.htm

http://www.harvest.com/spe.htm

I'm not sure whether they would depend on something proprietary at the
central end.


______________
regards,
Ron

In article <[email protected]>, John Dann
<[email protected]> wrote
>I'm pretty sure that there are (small) single box solutions (maybe
>many) already available, ie literally a GPRS modem with an external
>RS232 port and presumably a TCP/IP stack within the 'modem'.
[...]
>My take on this at present is that the companies that specialise in
>GPRS data links are making too much money from their existing
>customers to readily offer cheaper options. This isn't really a dig at
>them - a good business will make money wherever it can and nothing
>wrong with that. But at some point competition should kick in and
>progressively bring in lower cost solutions.

Perhaps if you were to indicate the price you would be willing to pay
for such a device and the volume involved then those with the ability to
do such work could form a judgement regarding the commercial viability.

--
Bob Evans

On Mon, 11 Jun 2007 14:28:18 +0100, Bob Evans
<[email protected]> wrote:

>Perhaps if you were to indicate the price you would be willing to pay
>for such a device and the volume involved then those with the ability to
>do such work could form a judgement regarding the commercial viability.

This is intended for a product extension that we would resell on an
individual basis not - at least in general - in batches. So it's
potentially more than a one-off but likely to be a succession of
individual units over time. So volume would be small but with
potential for limited numbers if the cost-benefit ratio were
worthwhile.

The unhelpful answer to the price question is as low as possible. A
trade price of £200-250 as initial purchase price would be a target,
preferably including commissioning the GPRS link. A monthly cost of
perhaps £20pm might be acceptable for passing maybe 200KB of data per
month. There might be some limited flexibility on these numbers but
not too much.

My main aim right now is to ty to ascertain whether this is feasible
(a) in principle and (b) within the cost envelope that this particular
market sector will bear. Then, if so, to work up a demonstration
system that can be used to get a clearer idea of market potential.

JGD

In article <[email protected]>,
[email protected] says...
> In article <[email protected]>, Ronnie <[email protected]> wrote:
> >On 11 Jun 2007 08:05:08 GMT, Gordon Henderson
> ><[email protected]> wrote:
> >
> >>These are still just modems - so the OP still needs that magic black box
> >>to take rs232 from his data equipment and interface it to a 2nd rs232
> >>link running PPP to talk via these modems...
> >>
> >A bit like one of those old-fashioned X25 PADs, except talking PPP
> >instead of X25. Actually, a bit more complicated than that, because
> >wouldn't this magic black box also need to do connection supervision
> >(connect, disconnect, re-connect, re-attempt)?
> >
> >Presumably the 'standard' asynchronous terminal servers expect a fully
> >functioning IP network behind them. So maybe a simple laptop running
> >linux could emulate that - present an IP interface on a network port,
> >and use a GPRS connection using some kind of PPP dialler? Plenty of
> >router distributions out there to do that - Freesco, for example, if
> >the GSM terminal has an RS232 port, or even Knoppix if the phone has a
> >USB connection. Still a 3, no, 4 box solution though - the OP's data
> >generating box, a terminal server, and a laptop - plus the GSM/GPRS
> >terminal.
>
> Quite - although you ought to be able to combine the termina server and
> laptop into one box - any Linux box with 'router' type functions and 2
> serial ports...
>
> Data Box <-- rs232/txt commands --> BlackBox <-- rs232/ppp --> GPRS modem
>
> If I had the time, I'd offer to build the blackBox for the OP, but I
> don't - I can give you a shopping list for about £200 of parts though,
> but the hard part will be getting a Linux to boot off flash, run the PPP
> session and provide some sort of protocol conversion to package up the
> data coming out of the DataBox and send it out via the Internet connection
> (and listen to the connection for commands coming the other way)
>

There are a fair few hackable devices which run linux and might serve as
the "black box". Things like the NSLU2 and various routers for example.
Some have one or more serial ports built in but not usually accessible
though there's always an RS232 USB adaptor. If the phone supports USB
that would be one less bit.

Steve.

In article <[email protected]>, John Dann
<[email protected]> wrote
>My main aim right now is to ty to ascertain whether this is feasible
>(a) in principle and (b) within the cost envelope that this particular
>market sector will bear. Then, if so, to work up a demonstration
>system that can be used to get a clearer idea of market potential.

My view is that what you propose is technically feasible and once
properly engineered could probably be supplied at a unit cost somewhere
near your target, assuming that the required housing were not too
elaborate. However margins would be very tight and so viability of the
whole thing would depend on having a design that goes together easily
and without burdensome setup costs that can make very small batch
production uneconomic.

Such considerations mean that the up-front engineering costs would be
significant and, not being absorbable within the unit price, would need
to be addressed separately.

On the positive side, I would expect the ongoing network costs for
moving the relatively small amount of data to be significantly below
your figure.

--
Bob Evans

Gordon Henderson wrote:

> These are still just modems - so the OP still needs that magic black box
> to take rs232 from his data equipment and interface it to a 2nd rs232
> link running PPP to talk via these modems...

Duh! Teach me to read.

--
Bruce Horrocks
Surrey
England
(bruce at scorecrow dot com)

John Dann wrote:

> The problem I've got, which prompted the original post, is that they
> seem to be part of customised high-cost solutions. But if they work
> for high-end solutions then there's perhaps no reason in principle why
> they shouldn't work for lower-end - ie lower-cost - solutions - the
> hardware cost isn't really the issue but it's the how-to knowledge and
> the provisioning of the GPRS service which seems to be costing.
>
> My take on this at present is that the companies that specialise in
> GPRS data links are making too much money from their existing
> customers to readily offer cheaper options. This isn't really a dig at

Have a look at <http://www.harvest.com/spe.htm>. Price for the base unit
is close to your range. Data will be on top, of course.

Regards,
--
Bruce Horrocks
Surrey
England
(bruce at scorecrow dot com)