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.

RTOS/CC1310: CC1310: how to get S/N ratio with TI-RTOS?

Part Number: CC1310

Tool/software: TI-RTOS

Hi,

I have an issue about radio communication on CC1310. When I transmit some packet (length is 580 byte, GFSK, boudrate 100kbps), CRC error was detected rarely.

To analyze this issue, I'd like to know S/N ratio at a receiver. How can I get S/N ratio with TI-RTOS?

I guess some interference affect to my rf, so I also would like to know rf settings that has resistance for interference.

Thank you for your kindly cooperation!

  • SNR is approx 7 for the PHY parameters you have used.

    I am not sure how you test, at what signal levels and at what range, but to put things into perspective: I assume you transmit packets. The packet error rate (PER) is related to bit error rate (BER) as follows: PER = 1 - (1- BER)^2n, where is the number of payload bits.

    For 10% PER (say) and 580 byte (= 580 x 8 bits) the BER is 0.0023%.
  • Hi Sverre,

    Thank you for your advise.
    If noise level at receiver on CC1310 can be measured, I can estimate affection of interference.
    Currently, my packet error rate is up to 0.5%, so it is difficult to evaluate using PER.

    If I want to more stability on RF, could I apply FEC? or any other solution?
  • One simple way of measuring the CC1310 noise figure (NF). Use SmartRF Studio, simple RX. Configure the receiver to a known RX filter bandwidth. Find the noise floor as the RSSI level with no signal at the antenna (terminate into a 50 ohm load). NF = noise floor + 174 -10log(RX filter BW).

    If you are concerned about interference you can also check if the noise floor is above the expected level. Sensitivity = Noise floor + SNR. For the 50 kbps test case in Studio the sensitivity (1% BER) is -110 dBm. Since sensitivity = noise floor + SNR you should measure a noise floor off approx -117 dBm.

    580 byte payload is rather long. You could consider reducing the payload to improve the PER. It will affect the throughput (more overheard), but might be something to consider.
  • Hi Sverre,

    I checked noise floor with SmartRF Studio, and the result is shown in above picture.

    In this case, NF = -90 + 174 - 10log196000 = 31. Is my calclation correct?

    you said

    > For the 50 kbps test case in Studio the sensitivity (1% BER) is -110 dBm. Since sensitivity = noise floor + SNR you should measure a noise floor off approx -117 dBm.

    how much the sufficient value of SNR?

    BTW, I consider to use FEC and DSSS to stabilize RF, so could you tell me how to use FEC and DSSS?

    for FEC or DSSS setting, have any RF parameters constraint?

  • The noise floor is too high in you measurements. This is probably due to interference. Try with a different RF frequency; 868 MHz say. Or do the test in a shielded environment. With RX filter BW set to 100 kHz you should measure a noise floor close to -117 dBm . With 200 kHz RX filter BW the noise floor will be 3 dB higher (approx -114 dBm).

    The SNR for GFSK with mod index = 1 (= 2 x deviation = symbol rate) is approx 7 dB.

    There will be support for DSSS + FEC in the next Studio release in July.
  • Hi,

    Yes, the noise floor in my measurement environment is too high, but I need to use this frequency.

    Currently, I try to monitor noise floor in the system. so could you tell me the command or function of TI-RTOS to know RSSI with no signal.
    I always use EasyLink_receiveAsync() to communicate on RF. I guess this function could not inform RSSI without valid packet.
    How to get RSSI with nosignal?

    it's my last question.
  • You can use the RF_getRssi function from the RF Driver once you have made sure that the radio is in RX mode.
    Please see
    dev.ti.com/.../_r_f_8h.html
    for more details.

    BR

    Siri
  • Hi both,

    Thank you for your strong support.
    I can get clear understanding of my isssue!