[AusNOG] 3G wireless signal Strength testing.

[David Connors]

On 9 June 2010 16:50, Thomason, Simon <Simon.Thomason at racq.com.au> wrote:

>  Was wondering if anyone has had a bit of experience with testing of
> signal strength of different carriers’ wireless 3G devices/services. Weather
> this be a piece of software or hardware that can provide a little more
> information that the standard fair from a USB dongle
>
> Ideally I am interested in a piece of software that resides on a remote
> machine that can report back to a head end that can keep historical
> information and can be reported on.
>
I've done a fair amount of this stuff before. IMEI define standard
AT-commands you can use to control/query the mobile terminal. The IMEI specs
are *very* good and the Falcom and Nokia devices I've used in the past have
very consistent behaviour (i.e. follow the spec well).

In Windows, and depending on the device, it will normally present to you as
a number of asynchronous serial devices as COMn: ports. One of these will be
the command device that lets you control the mobile terminate via AT
commands (refer to the IMEI specs for this - they are really good). I was
going to tell you that that is probably no use to you when your vehicles are
on the road because that port is probably locked but I just tested it on my
lappy and the command port is fine for use when the Sierra stack is
connected and working.

Hence:
at+csq
+CSQ: 20,99

The first value is the receive signal strength indicator (RSSI) and the
second is the bit error rate. 99 = unsupported on this modem. No matter, you
don't need the BER value anyway. There are tables that convert CSQ values to
dBm but you probably want to read the manufacturer specs for your MT gear to
work out what is available and what special cases/notes there are. I have
not looked into the specifics of getting CSQ figures for different bands and
so on. There might be a revised set of commands that exposes that stuff.

So in practical terms, and I assume your platform is those in-vehicle
windows boxes on the RACQ trucks that bring me batteries, I'd probably
suggest that you write an NT/CE service that runs on your fleet vehicles
that:

   1. Grabs the current RSSI from the MT command device using AT commands.
   2. Grabs the current lat/lon pairs from your onboard GPS.
   3. Log the above into a SQLExpress/Compact instance on the vehicles.
   4. Punts that up to a head end web service periodically.
   5. Presto, you're done.

While you're at it and seeing as you're on the road, check unencrypted
802.11abgn traffic on 2.4 and 5GHz, maybe addresses of famous peoples
nearby, and any salacious web sites and grab them too for serving ads
later.

Also, and more seriously, I would fortify any sort of dBm figures you
collect with at least a perfunctory real-world performance/transfer test.
After all, the reported signal strength values can be seen as advisory to
some degree - but mbps actually copied to your remote terminal - well, that
is *real* and hard to fudge. :)

> Also was wondering if anyone could give an idea of there experience with
> latency over different networks in regards to 3G. A good comparison would be
> the difference that can be experience from a wired connected to a 3G in the
> same area to the same destination.
>
No one probably outside of RACQ can give you those A/B figures as it is very
difficult to do for a corporate and not entirely fair in some cases. The
Internet side of the IP traffic may emerge at different places at different
times on the terrestrial network depending on what the mobile operator wants
to do. Also, for a corporate like RACQ, you generally get special SIMs with
special certs/APNs/etc that terminate the traffic inside your own corporate
network via either a VPN or other private link with your own internal
addressing and route to the Internet. In that scenario, you have a very
special and unique scenario for egress from the wireless network that no one
outside your company can really comment on or help you with in terms of
generating a like-for-like comparison. Sorry.

Unscientifically however, and IMO, there is only one game in town for
wireless and that is NextG. I persisted with a 3 dongle for some time but
any time I went even remotely out of town there was nothing but GPR$ to fall
back to. That might have improved recently - but I doubt it. My informal
testing on other people's iPhones on Voda and Optus has been pitiful by
comparison.

Also, the *type* of 3G connection has a big impact on latency. HSDPA vs
HSUPA etc.

I hope that helps.

David.

-- 
David Connors (david at codify.com)
Software Engineer
Codify Pty Ltd - www.codify.com
Phone: +61 (7) 3210 6268 | Facsimile: +61 (7) 3210 6269 | Mobile: +61 417
189 363
V-Card: https://www.codify.com/cards/davidconnors
Address Info: https://www.codify.com/contact
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ausnog.net/pipermail/ausnog/attachments/20100609/cd23f26b/attachment.html>