Re: [dispatch] SIP and GSM/UMTS with OpenBTS

[Michael Hammer <[e-mail]>]

Interesting.  But, the 4G/LTE network is an IP internet of sorts.
The SIP/IMS components do not need to be in the radio network, 
and thus not at the locations where the power is constrained.

But, I'll mind my own suggestion and not go there.

Michael Hammer
Principal Engineer
[e-mail]
Mobile: [phone]
500 Yosemite Drive Suite 120
Milpitas, CA 95035 USA


-----Original Message-----
From: Jim Forster [mailto:[e-mail]] 
Sent: Wednesday, February 05, 2014 8:46 PM
To: Mary Barnes
Cc: Michael Hammer; [e-mail]
Subject: Re: [dispatch] SIP and GSM/UMTS with OpenBTS

Mary, all,

Lots to read and consider.  I'll be offline for 12 hours as we drive to
Delhi.

> I don't necessarily think 3GPP is out of scope in that I think it's very
important that if IETF decides that we will be documenting or even defining
protocol elements to support OpenBTS, it needs to be clear why 3GPP Call
control/SIP specifications aren't being used. 

Umm, one possible reason but I need to read and think, would be the question
of whether 3GPP Call control/SIP is in some ways tied to 4G/LTE in some way.
OpenBTS puts 2G/3G phones on the Internet.

> In offline in discussions with the proponents, it's my understanding that
OpenBTS is being used in very specific environments where there is no full
MSC deployments (i.e., remote locations that have no deployed IMS networks).
The intent isn't to replicate a full IMS system, but rather to provide basic
connectivity to the Internet and use SIP in a non-IMS context to complete
the calls.   So, I think more detail about this context would be very
helpful.

Yes, that's correct.  No full MSC; in fact only Asterix or Freeswitch or
Yate, usually embedded in the same device as the RF and signaling.

The goal was and is to make really small, low power, devices that can be
deployed where classic systems don't fit, usually because they cost too much
and take too much power.  Small islands, Antarctic research stations, the
jungle of Paupua Indonesia, etc. 

This very reduced system, which might possibly be called IMS Lite for 2G/3G
phones, seems to be very useful to people that want to make calls and send
SMS in those places.  It's not clear (largely because of my ignorance) how a
full IMS benefits those scenarios, while the simple OpenBTS
translation/gateway of 2G & 3G calls to SIP is a package that people want.
Well, to be clear, people with phones don't care what protocols are used,
they care about other aspects, like power and cost.

> I think it would really help if there were some diagrams showing what IMS
protocols are being used.

Well, I can certainly provide some diagrams that show how the system works.

>  My understanding is that current implementations have various ways in
which they are interworking the Radio Layer Call Control messages to SIP
messages.  I think the motivation is to improve interop by defining a
consistent "mapping" if you will.

Correct.

>  It's not clear to me how exactly registration is being handled - perhaps
mapping some of the Radio Layer mobility management messages.   My
understanding is that there can be no change to the existing messages from
the cell phone to the BTS for obvious reasons, so all the interworking needs
to happen in the BTS.  

Correct.

  -- Jim