RTOS/TMS320C6748: VoLib CID component - Channel Seizure Lengh too long for BT standard

Hi Jonathan,

I've forwarded your query to the software experts. Their feedback should be posted here.

BR
Tsvetolin Shulev

Hi,

If you already set the leading_marks = 0, there is no further way to reduce the detection time. If you got 170 ms, isn't it still within the range of BT spec., (80 ms < Channel Seizure Duration < 262 ms)?

Regards, Eric

Hi Eric,

For example, the CO may send a CallerID with Channel Seizure Length = 100ms  which is still within the BT specification. In this case, the TI VoLib CID component cannot detect this CallerID.

It is possible that this is just a constant in the source code.  Is it possible to modify this value in the source? It is possibly in the function void cidDtctFsmUsEtsi(void *cidInst, tint bit).

Thanks,

Jonathan

 

Hi,

The typical channel seizure signal consists of 300 continuous bits (duration of 250 ms at 1200 bps) of alternating “1″ and “0.” So a detection within 170ms will work for this case. But it may miss the shorter signal of ~100 ms.

It is possible that we have some setting for the duration of the signal to avoid false detection. However, the VOLIB has only posted on-line with library format and it is in maintenance mode for years. We don't have plan to modify the source code for any new library. Sorry for this!

Regards, Eric

Hi Eric,

Fair enough.   

I think I have a solution. As the CWCID does not has channel seizure signal, I think I can enable the call_waiting flag in the CID component. So I don't need to worry about the short channel seizure signal issue because the detection routine will skip it.

The draw back is the potential false detection. Maybe I can detect the CAS tone to trigger the CID detection to minimize this. But then, I went into another problem. I cannot make the CID CAS tone detection work.

Do you mind helping me to find out if the TDU can detect the CAS tone? In the TDU example, the macro tdu_SIM_USE_CIDCAS is commented out. Even I uncomment that, It still does not work.

Any idea?

Thanks,

Jonathan

 

Hi,

From the \ti\mas\tdu\docs\doxygen\TDUAPI.chm, standard compliance -----> primary detector details. CID CAS 2130 Hz & 2750 Hz Dual Tone Detector is supported.

Regards, Eric

Hi Eric,

I understand that CAS tone detection is part of the TDU. However, I cannot make it works with the example (tdusim.c). This is how I try to test that.

I create a u-law CAS tone files with 8kHz sampling rate. Add the following line before all the detection. Set the breakpoint at the detection callback.

stfAllDetectionFile(tduInst, "cas_axs.pcm");


Nothing happens


Then, I makes two changes and makes no difference

Change
#undef tdu_SIM_USE_CIDCAS to #define tud_SIM_USE_CIDCAS

Add
&tduIncCIDCAS to the tduDetContext


Maybe I have missed something?


Thanks,



Jonathan

Hi,

There is input PCM example type defined in the tdusim.c (a-law, u-law or linear). Do your code match the commanding of the PCM file you created? Also, there are many test input vectors in the package, did those work for you for various tone detection?

Regards, Eric

Hi Eric,

All the test input  vectors in the package works as it is used in the example. All these vectors are in u-law as specified the the example. I also created a 1004Hz Test Tone in u-law. Also works without problem. But I have no luck with the CID CAS tone (2130 Hz & 2750 Hz Dual Tone).

I have attached the u-law CAS tone file for your reference.

Thanks,

Jonathancas_axs.zip