Performance in German Vodafone 3G network

Trying to set up a ppp mobile broadband connection within the German 3G
network (APN mail.partner.de) I experience serious performance issues:
The connection lives up to some 300kb/s but then drops down to a few kb/s
for extended time. The modem (Samsung Z100) claims reception to be good
and 3G/HSDPA to be available.
With the very same settings (different APN, of course) and hardware
connections with the Telefónica mobile network are quite nice and stable.
So for the time being I prefer to pay Telefónica though Vodafone is
included in my ISP contract and basically free of charge.
Can anybody confirm that Vodafone Germany simply sucks? Any suggestions
on how to tweak the connection settings?

My wvdial.conf:

[Dialer Defaults]
Phone =
Username =
Password =
New PPPD = yes

[Dialer vodafone-usb]
Init1 = ATZ
Init2 = ATQ0 V1 E1 S0=0 &C1 &D2 +FCLASS=0
Init3 = AT+CGDCONT=1,"IP","mail.partner.de","",0,0
Init8 = ATM1
Modem Type = USB Modem
Modem = /dev/ttyACM0
ISDN = 0
Phone = *99#
Password = ich
Username = geheim

Günther

Marc Haber wrote:

> Günther Schwarz <(E-Mail Removed)> wrote:
>>Can anybody confirm that Vodafone Germany simply sucks? Any suggestions
>>on how to tweak the connection settings?
>
> Mobile network performance is highly dependent on your location.
> Vodafone is generally slightly better than O2, but there are places
> where the contrary is the case (for example, my apartment). But in
> difficult situations, such as in a moving train, I'd stick with Vodafone
> any time.

Moving trains are a different matter. Reception with Telefónica is quite
often simply limited to train stations. But then for the time being I'm
just sitting on my couch.
I will give it a try in a different location. So mobile broadband means
that I will have to get some exercise and take a walk in order to read my
mail? That gives the concept a new meaning ;-)

Günther

Marc Haber wrote:

> Günther Schwarz <(E-Mail Removed)> wrote:
>>I tried the office: no way, but that is all steel and concrete. Right
>>now I'm sitting outside an Italian bar right in the middle of a big
>>metropolian area. Throughput is in the kb range. OK for nntp, but not
>>for much else.
>
> What kind of Hardware?

SAMSUNG SGH-Z630 connected via USB or Bluetooth to a Thinkpad X300. I
have some USB stick from ZTE (?) also but do not know how to set this up
with Debian.

> Are you sure that you're connected via UMTS?

How can I tell? The phone manages the connection to the network. It's
display shows the symbol for HSDPA and almost maximal signal strength.

> How is the latency to a pingable host on the Internet?

Today it works better with 100-300 kb/s for a ssh copy from a reliable
host with good bandwith. Results for a ping:

$ ping www.heise.de
PING www.heise.de (193.99.144.85) 56(84) bytes of data.
64 bytes from www.heise.de (193.99.144.85): icmp_seq=1 ttl=245 time=447 ms
64 bytes from www.heise.de (193.99.144.85): icmp_seq=2 ttl=245 time=383 ms
64 bytes from www.heise.de (193.99.144.85): icmp_seq=3 ttl=245 time=400 ms
^C64 bytes from www.heise.de (193.99.144.85): icmp_seq=4 ttl=245 time=395
ms

--- www.heise.de ping statistics ---
4 packets transmitted, 4 received, 0% packet loss, time 3012ms
rtt min/avg/max/mdev = 383.639/406.789/447.562/24.315 ms

Just for comparison the results for my landline connection (access to the
router via WLAN):
Ping times are reduced by a factor of ten and speed is in the range where
HSDPA claims to be (though still less than what I pay my telco for).

Günther

Marc Haber wrote:

> Günther Schwarz <(E-Mail Removed)> wrote:

>>Today it works better with 100-300 kb/s for a ssh copy from a reliable
>>host with good bandwith. Results for a ping:
>>
>>$ ping www.heise.de
>>PING www.heise.de (193.99.144.85) 56(84) bytes of data. 64 bytes from
>>www.heise.de (193.99.144.85): icmp_seq=1 ttl=245 time=447 ms 64 bytes
>>from www.heise.de (193.99.144.85): icmp_seq=2 ttl=245 time=383 ms 64
>>bytes from www.heise.de (193.99.144.85): icmp_seq=3 ttl=245 time=400 ms
>>^C64 bytes from www.heise.de (193.99.144.85): icmp_seq=4 ttl=245
>>time=395 ms
>
> That's a typical EDGE latency. HSDPA should be in the 100 ms range.

Performance is indeed comparable to another SIM-card on a location where
no UMTS is available. So is Samsung to blame for the invention of
cheating electronics? ;-)

>>Ping times are reduced by a factor of ten and speed is in the range
>>where HSDPA claims to be (though still less than what I pay my telco
>>for).
>
> It's a common misconception that you pay your telco for 16 Mbit just
> because the product you bought contains the string 16000. Read the fine
> print and get acquainted with DSL physics, which generally ignores
> marketing..

That is true an well said. But I still do not get what the router
appliance tells me about the setup of the line. This gets OT. As far as
networking is concerned appliances take over for the sake of convenience,
but less control is the price one has to pay for that.

Günther

Warning:

Lew Pitcher, who posts to this newsgroup, is a domain thief.

Read the full story at http://www.lewpitcher.ca

> >$ ping www.heise.de
> >PING www.heise.de (193.99.144.85) 56(84) bytes of data.
> >64 bytes from www.heise.de (193.99.144.85): icmp_seq=1 ttl=245 time=447 ms
> >64 bytes from www.heise.de (193.99.144.85): icmp_seq=2 ttl=245 time=383 ms
> >64 bytes from www.heise.de (193.99.144.85): icmp_seq=3 ttl=245 time=400 ms
> >^C64 bytes from www.heise.de (193.99.144.85): icmp_seq=4 ttl=245 time=395
> >ms

> That's a typical EDGE latency. HSDPA should be in the 100 ms range.

Wow - that is also getting close to the round-trip latency one would
see using a GEO satellite both ways - O(500ms).

rick jones
--
It is not a question of half full or empty - the glass has a leak.
The real question is "Can it be patched?"
these opinions are mine, all mine; HP might not want them anyway...
feel free to post, OR email to rick.jones2 in hp.com but NOT BOTH...