Half-rate data calls (CSD)

[Vadim Yanitskiy]

Hi Mychaela,

(I changed the Subject when replying, since this topic is not specific 
to Sony Ericsson K2xx.)

On 23.09.2024 04:58, Mychaela Falconia wrote:
> Speaking of CSD, the one readily apparent limitation of CSD support in
> TI's full solution is the lack of support for CSD on TCH/H.  More
> precisely, I see various definitions for CSD on TCH/H, it looks like
> certain parts are present, but the end-to-end solution clearly lacks
> working support for these modes.  I haven't studied it enough to tell
> if the limitation is in the DSP ROM (unalterable), in TI's official
> DSP patches (amenable to modification in theory, but extremely
> difficult) or in ARM firmware which could be easily extended.  Has
> Vadim or anyone else looked into it?

I would be interested to play with the DSP directly using osmocom-bb 
code base.  While working on adding CSD support for it, I had to make 
only a small change to the layer1 firmware to get the channel coding 
working:

https://cgit.osmocom.org/osmocom-bb/commit/?id=069b627abecad7efb4de43e3bad7ceb685d65e89

Back then I was primarily focusing on getting full-rate data calls 
working and sadly had no time to test half-rate calls.

In theory, both TCH/H2.4 and TCH/H4.8 should be supported by 
osmo-bts-trx;  we even have TTCN-3 testcases for that.  There might be 
missing bits in osmo-bsc and/or osmo-msc, though.

> For Vadim or anyone else who may wish to look into it, look in L1 code
> and particularly in L1-DSP interface for something called IDS - that
> specific C preprocessor symbol.  (It presumably stands for Integrated
> Data Services, IIRC.)  It looks like the DSP has some dedicated words
> or buffers in its shared memory interface that are devoted to this IDS
> feature, but I haven't taken the time to study exactly what they
> correspond to in the air interface layer stacking.

If I remember correctly, it was mentioned somewhere (or by somebody in a 
conversation) that TI's DSP has a hardware accelerated module for the 
V.110 stuff (RA1'/RA1/RA2 functions).  I guess this could be it?  In the 
case of osmocom-bb, we're doing all the bit manipulations in software on 
the host side.  This is not a big deal for a modern x86_64 machine, but 
could likely be offloaded to the DSP in the phone's normal firmware.

-- 
Best regards,
Vadim.