• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 2 answers
  • Subscribers 30 subscribers
  • Views 622 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

IWR6843: How much loss due to CFAR?

Johanwww
Intellectual 500 points
Part Number: IWR6843
Hi

We are making a system analysis and need to know the losses due to the CFAR algorithm.

I understand this is a very broad question which depends on many things tough.

Do you have any knowledge to share on how this is calculated based on window size for noise estimation, probability of false detection and probability of detection?

We are using CFAR_CASO in range and in doppler CFAR_CA, if that can make  this any easier.

cfarCfg -1 0 2 8 4 3 0 1200 0
cfarCfg -1 1 0 4 2 3 1 1200 0

  • 0
    TI__Mastermind 35645 points

    HI, 

    Given SNR, given the window size, the probability of false detection and detection rate is fixed, you can find in following reference:G. Dillard, "Mean-Level Detection of Nonfluctuating Targets", IEEE AES-10, 1974.    Of course with same detection threshold (SNR), the larger the window size, you will get smaller false detection.   

    Different CFAR variation like CFAR-CA, CFAR-CASO, is a variation in calculation SNR.   In single target case, CFAR-CA the standard method works well.   Then if there are multiple targets in the similar range bins, CFAR-CASO can help to measure the noise with less impact on from the other target.

    Hope it helps,

    Zigang

        

  • 0 in reply to zigang Yang
    Intellectual 500 points

    Hi,

    Have you made any analysis or do you have any data on the losses due to the CFAR algorithm?

    Kind regards Johan

  • 0 in reply to Johanwww
    TI__Mastermind 35645 points

    Yes, here is the table for SNR needed to achieve different level of false probability rate.  You can find it from the reference book I sent you previously.

    The column is the false probability rate, and each row is average window size.

    AvgLen\FalseProbRate          

    10-8

    10-7

    10-6

    10-5

    10-4

    10-3

    10-2

    8

    72.00

    51.99

    36.99

    25.74

    17.30

    10.97

    6.23

    16

    34.60

    27.81

    21.94

    16.86

    12.45

    8.64

    5.34

    24

    27.71

    22.98

    18.68

    14.77

    11.23

    8.00

    5.08

    32

    24.90

    20.95

    17.28

    13.86

    10.67

    7.71

    4.95

    40

    23.40

    19.85

    16.50

    13.34

    10.36

    7.54

    4.88

    48

    22.45

    19.15

    16.01

    13.01

    10.15

    7.43

    4.83

    56

    21.81

    18.68

    15.67

    12.78

    10.01

    7.35

    4.80

    64

    21.35

    18.33

    15.42

    12.61

    9.91

    7.29

    4.77