• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 9 subscribers
  • Views 4242 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.

Mathematical method to calculate the best Sync word for CC1101 WOR as 0xD391

TYB
Prodigy 165 points
Other Parts Discussed in Thread: CC1101

Hello TI expert,

I'm using CC1101 for my RF application in star topology with about 250 child nodes, all WOR enabled.

I'm interested in waking up child nodes selectively. For that I'm using different sync words for each node. According to the TI resources I have to take care when I'm selecting a sync word so that I won't get false alarms. I've read that I need to be concerned about the autocorrelation of the sync words. Could you please explain how to select the best sync word set?

Further, resources indicate that the optimal sync word would be 0xD391. It would be very helpful to understand how this particular sync word was selected (mathematically). Your invaluable time spent on helping me to understand this is thankfully appreciated.

Thilina

(Sri Lanka)

  • 0
    TI__Guru**** 317180 points
    -Convert the sync word to a binary vector
    -Run xcorr(2*sync_vector - 1) in Matlab or equivalent
    -Plot the result
    •A good sync word has no side lobes
    For this application I would also look at the address filtering built in in CC1101. The node has to be up in Rx for a given time. Use address filtering and only wake the MCU up if the correct address is received.
  • 0 in reply to TER
    Prodigy 165 points

    Thank you very much TER for the very quick answer,

    I just checked with Matlab and understood that for the sync word 0xD391 we get a decent autocorrelation plot.

    Giving different addresses is a little out of focus in my project and I'm using my own protocol for addressing. Now I understand that the particular sync word is optimum, could you please shed some light on how to get that value (0xD391). If I want to calculate the next best 100 sync words, what would be the best method for me to follow.

    Thanks again.

    Thilina

  • 0 in reply to TYB
    TI__Guru**** 317180 points

    I suspect that it will be difficult to find 100+ sync words with decent auto correlation. We don't have any methodology to find sync word, just how to test them. In addition to the auto correlation the sync word has to have enough transitions and be close to DC balance. So even if you have your own addressing using the address filtering in the chip instead of the sync word to wake up the node will most probably be a faster and more robust way to implement the system.

    In addition you have to take into account that if you use a 16 bit sync word you will get false syncs at time since it is a given probability to find a 16 bit sync word in noise. If you use 32 bit 1 bit error is allowed for each 16 bit word which could also cause some issues if you need 100+ different sync words.

  • 0 in reply to TER
    Prodigy 80 points

    Hello experts, 

    I've read a little bit more on selection of proper sync words for this kind of an application. According to the TER, the autocorrelation plot of the recommended sync word (0xd391) would be as follows.

    It has lower sidelobes with a decent dc power. I've read about Gold codes, m-sequence, walsh-hadamard and 0xd391 doesn't seem to be calculated from any. I wonder this particular recommended sync word is selected based on certain other parameters also. Any insight which will allow me to research further is deeply appreciated.

    Thank you for your valuable time.

    Thilina

  • 0 in reply to Thilina Yapa Bandara
    TI__Guru**** 317180 points

    - Good auto correlation

    - DC balance

    - Most possible transitions '1' <-> '0'

    The sync word was decided for many years ago so I'm not able to find out how exactly thay found this sync word.

  • 0 in reply to TER
    Prodigy 165 points

    Thank you TER for your very valuable insights in this forum. The last reply is marked as solved.