AWR1642: Question on CFAR

Tianyuan Yang

Prodigy 30 points

Part Number: AWR1642

Hi:

In TI's mmW Demo,I found that CFAR is implemented in logarithmic scale，

Function：

mmwavelib_cfarCadBwrap（）；mmwavelib_cfarCadB_SOGO（）；

What is the benefit of Ti choosing to complete CFAR in logarithmic scale rather than real abs value ?

Thanks

over 7 years ago

0 Akash Gondalia over 7 years ago

TI__Guru 50240 points

Hi Tianyuan,

We have received your question and it has been assigned to a mmWave team member. Someone should have an answer for you shortly.

Cheers,
Akash

0 Anil Mani over 7 years ago

TI__Expert 4720 points

HI Tianyuan,

The C674x DSP can efficiently convert complex numbers to the logarithm of their absolute value. Also, converting a number to a 16-bit logarithm results in a space saving (from 32 bits per complex sample to 16 bits per logarithm of sample), without incurring a reduction in dynamic range.

There were no other reasons to select log-CFAR over linear-CFAR in the SDK. Other approaches have also been tried. . The 'traffic monitoring demo' () for instance, uses linear-CFAR.

Regards

Anil

0 Tianyuan Yang over 7 years ago in reply to Akash Gondalia

Prodigy 30 points

Thank you！ :)
Regards
Tianyuan

0 Tianyuan Yang over 7 years ago in reply to Anil Mani

Prodigy 30 points

Thanks you,Anil :)

Regards
Tianyuan

Sensors

Sensors forum

AWR1642: Question on CFAR