Slashdot Mirror

MMS: Multimedia Messaging System -- this allows you to send email with audio/amr attachments, so you can play them with open source code from public 3GPP technical standards TS 26.071, TS 26.073, TS 26.101 and TS 26.074.

AMR is a truly great vocodec technology, which stands for "Adaptive Multi Rate." Accordingly, it takes less bandwidth and battery time because when the microphone isn't picking up sound (from, e.g., your speech) your phone isn't sending as much information. Just the way it ought to be. Why spend 4 kbps to send comfort noise when 20 bps can do just as well? Execellent code!

This is true, but after you have gotten an IP to your device, this IP is what you look after to charge you. If you can change it (and the router back to you offcourse) you wouldn't be charged, would you?

Actually, yes you would be charged. It's not the IP-address per se that the network looks as in a 3G network to decide who sent (or received) how many bytes to whom (or was active for a certain period of time, 3G allows both), but the tunnel ID.

You see, all end user traffic in a 3G core network (which does the charging part) is tunnled over a protocol called GTP, each user (i.e. active PDP-context of each user and QoS level) has it's own tunnel. The network never really looks at the end user traffic, it just switches tunnels. So in effect, changing your IP address would only prevent your IP stack at the mobile/laptop from accepting the packets, not the network from actually delivering and charging you for them. (Assuming PDP-type-IP).

This is the way it must work if the operator is to be able to correctly isolate corporate customers, without any overlap with other customers. Corporates, that is that may use private addresses and NAT to connect to the Internet per se. So, in effect your phone may not be the only one in the network with that very IP address.

Now, IPv6 complicates matters some, but not much, the basic IPv4 3G infrastructure is still there.

If you want to know more about these matters, it's no longer a secret. All the 3G specs can be found at 3GPP. Start with the 23.060 specification, it's the overview. From there on you can dvelve deeper into the charging and the GTP specs, though they are not for the faint of heart (and heavy to carry around to).

Corrections and Extra Info

What they are actually referring to is "CDMA 2000-1XEV". The "EV" stands for "Enhanced Voice".

VZW is rolling 1xRTT, 1xEVxx stands for EVolutionary

WCDMA: Wideband CDMA. It wil start to appear in Europe and some US networks later this year (IIRC). It will have a maximum data rate of 384Kbps (IIRC). However, it uses almost 5MHz of the spectrum (~2.5 forward link + ~2.5 on the reverse link)

Maximum data rate is NOT 384 kbps, this is just what the Japanese early FOMA adopters are limited to. The 5MHz is NOT split forward/reverse. Bandwidth and chip rate (3.84 Mcps) is same in BOTH directions.

1xEV-DO stands for EVolutionary Data Only

For those with even a slight interest in the actual facts and standards, there are two standards groups looking after the two 3G streams. The 3G Paternership Project is responsible for the GSM migration path aka WCDMA. The 3G Partnership Project 2 covers the Qualcomm migration to cdma2000 (1xRTT, 3x, etc) etc.

The current dominant standard, GSM, is evolving toward WCDMA/UMTS. This is the upgrade path all GSM and most IS-136 networks will take. GSM/GPRS uses the same data infrastructure and protocols as WCDMA/UMTS. Oddly enough, GSM is more different from WCDMA/UMTS than IS-95 (CDMA) is from CDMA2000. CDMA also has some technology advantage over GSM. But it's kind of a Betamax thing: it is better to be widely used than it is to be better technologically.

Both 3G standards use CDMA technology, but they are not compatible. Maybe there will be dual mode radios that will be cheap enough to work on both kinds of networks.

Anybody know what iDen's upgrade path is?

How do I turn off the redundant things in the square brackets?

If this network is sucessfully deployed (think 2002 to 2004), it should give IPv6 a huge shot in the arm.

Scientific American discussed this last October.

The Third-Generation Partnership Project is the group developing the standard.

For more information on what is going on in the standards arena, you may want to check out 3GPP at http://www.3gpp.org.
---

It all comes down to the old chestnut that IP is a best effort service, and the need to find a scalable solution that allows the reservation of bandwidth (Diffserv is scalable, but does not allow bandwidth reservation, only relative QoS, wheras RSVP allows bandwidth reservation but does not scale well).

regards, treefrog

As someone who works with the standardization of 3rd generation mobile systems (specifically I am a 3GPP delegate [3GPP website]), I can tell you that those speeds will only be available for indoor type connections, like inside the lobby of a hotel or the waiting area in a train station. The speeds decrease as you move to pedestrian, moving car, train, etc.
---

There is a problem with getting a radio to work on multiple frequencies, but soft radio is on its way (i.e. a software programmable radio receiver).

And as for open source - well, there is an awful lot of *very* proprietry (Read patented) stuff out there - licenecing of patented ideas is a major revenue stream for the equipment manufacturers. However, the standards are open, and (for UMTS - the 3G system) held here .

You are right about one thing though; things are different in the US because their frequency allocation is diffrent from the rest of the world. Their cellular standards are different (e.g. IS95), but there is an awful lot of cross-fertilization of ideas between IS95 and UMTS, because the manufacturers are looking to minimise their costs.

regards, treefrog

The usage of data services (and hence the revenue) is only 1% even in Japan. So any solution has got to pay for itself on something other than data. Maby in 10 years time we will have 50% of usage being data but for the vast majority that isn't happening yet.

Basically, the network is composed by a set of cells that (ideally) cover the area where costumers are. Those cells are served by Base Stations. The Base Stations have antennas, and those antennas, being non ideal, have a certain directivity.

As most users are on the ground, the antennas are designed to have their best response in the direction pointing from the Base Station to the ground (usually -3 to -10 degrees).

On the other hand, the antennas have the worst directivity in the areas that there are less users (you don't want to waste your power). So antennas do not send almost any power towars the air (or the ground below for the same reason).

That is mainly why you can't use a mobile phone in a (flying plane).

As for the data rates, they are in MegaBITS per second (Mbps), not MegaBYTES per second. And those claims are just theoretical.

All 3G (3rd generation) proposed architectures claim speeds up to 2Mbps when the user is still or walking at very low speed, and up to 384 Kbps at velocitis of up to 500 Km/h! (This is so that you could use your phone on a (future) high speed train to connect to the Internet).

But those are theoretical fields, and (almost)everybody working in this field (as I do) believes that the real data rates will be of 384 Kbps when STILL and about 114 Kbps when going at high speed (but probably not that high!).

About the frequency plan that somebody asked about I recomend to check the 3GPP and the IMT-2000 sites, where you'll find tons of info about 3rd generation mobile phones.

And, by the way, Qualcomms proposal is only one among many others, although most of them are based on Wideband CDMA technology, and the only comercial(narrowband) CDMA network is Qualcomm's IS-95, which is mainly deploid in the US, Brazil and some oriental countries. Qualcomm's CDMA proposal is called (last time I checked!) cdma2000, although the extension of IS-95 is cdmaOne, and I don't know if the one they are talking about in this article is the first or the latter.

I hope this clarified some of the issues raised here!

Cheers!

Angel