locked
Performance of mobile test with SORA RRS feed

  • Question

  • Hello!! I want to measure network performance of WiFi with SORA in mobile situation (with car). However, as I know, SORA parameters (ex: rxgain, txgain...) should manually set though optimal values are different according to situations(ex: distance between two devices, device, noise). So, would you give me any advice how I can support mobile test with SORA? Thanks.

    Saturday, September 3, 2011 8:28 PM

Answers

  • Hi Jaeims,

    It's an interesting idea to test Sora in mobile setting, and your concern is valid. However, we don't have such experience and may not be able to provide suggestions. I think, at least you need to add an AGC (Automatic Gain Control) into current WiFi implemetation.

    -Jiansong

    Tuesday, September 6, 2011 2:37 PM
  • valid values:

    rxpa: 0x1000, 0x2000, 0x3000

    rxgain: 0x000, 0x200, ..., 0x3E00

    txgain: 0x00, 0x80, ..., 0x1F80

    Wednesday, September 7, 2011 6:22 AM
  • Hi Jaeims,

    There is a transmission power limit of the RF board. Please refer to your RF board data sheet to find out it, e.g. USRP XCVR2450.

    The sharp cut in the delivery ratio (i.e. you find a clear transmission range) may also be caused by the threshold setting in the receiver baseband. You can try to make a few dumps and use software oscilloscope to examine the details.

    Thanks,

    - Kun

    Thursday, September 8, 2011 6:16 AM

All replies

  • Hi Jaeims,

    It's an interesting idea to test Sora in mobile setting, and your concern is valid. However, we don't have such experience and may not be able to provide suggestions. I think, at least you need to add an AGC (Automatic Gain Control) into current WiFi implemetation.

    -Jiansong

    Tuesday, September 6, 2011 2:37 PM
  • Thank you for response.

    For gain control, I have to know the effect of rxgain and rxpa as well as txgain.

    I know from previous article in FORUM that rxpa have its value 0x1000, 0x2000, 0x3000 and rxgain have its value 0x000~0x1F000. Then, how about txgain? Is there any unit for setting txgain?

    Thanks

    Wednesday, September 7, 2011 3:24 AM
  • valid values:

    rxpa: 0x1000, 0x2000, 0x3000

    rxgain: 0x000, 0x200, ..., 0x3E00

    txgain: 0x00, 0x80, ..., 0x1F80

    Wednesday, September 7, 2011 6:22 AM
  • Thank you for your answer.

    I have question again. Is there any way to increase txgain? I set txgain to 0x1F80, which is maximum value you mentioned in reply. However, i found that if distance between two device is long (for example 10m), I found that it is prone to noise or interference.

    I think that only increase rxgain or rxpa is not much helpful since increase of rxgain amplifies WiFi signal as well as noise signal. Moreover,  for long distance case, data signal is weak. That means SNR is low, so performance may not be stable only by small change of situation. (for example, posture of antenna, small increase of noise). So, I think that increase of txgain should be necessary for communication of longer distance. So, I want to know how I can increase txgain.

    In addition, I want to know what is the maximum distance you guys experiment between two Sora devices. In my case, until 7m, communication works well with default 802.11b reference code which you distributed. However, if I extend transmission range larger than 10m, Icommunication between two Sora devices failed with various possible combination of rxgain and rxpa.

    Thanks.


    • Edited by Jaeims Wednesday, September 7, 2011 6:51 PM
    Wednesday, September 7, 2011 7:43 AM
  • Hi Jaeims,

    There is a transmission power limit of the RF board. Please refer to your RF board data sheet to find out it, e.g. USRP XCVR2450.

    The sharp cut in the delivery ratio (i.e. you find a clear transmission range) may also be caused by the threshold setting in the receiver baseband. You can try to make a few dumps and use software oscilloscope to examine the details.

    Thanks,

    - Kun

    Thursday, September 8, 2011 6:16 AM
  • Thank you for your reply.

    I use dauterboard as XCVR2450, whose typical tx power is 100mw.

    I found that packet loss is very low (around 0%) when the distance between two device is 5m by adjusting rxgain and rxpa. However, when I touch antenna a little bit, (just touch, not move), transmission is not successful (packet loss is almost 100%). Of course, I use default 802.11b code.

    However, when transmission range is short, (around 1m), packet loss between two SORA devices is always almost 0%. (not influenced by test environment) So, what I guess that from this pheonomenon, reception of packet is prone to interference and noise when communication range is long. This is because of received signal power is low when distance is long. So, I want to test transmission by increase tx power by adjusting txgain. However, according to above answer, maximum txgain is 0x1F80, which I use for txgain in this experiment. (and also default value for txgain.)

    So, my question is

    1) Is it possible to increase tx power by adjusting txgain? (my txgain setting is 0x1F80)

    2) Is it possible to extend communication range to 30m just by adjusting rxgain and rxpa carefully?

    Thank you.


    • Edited by Jaeims Monday, September 12, 2011 11:36 PM
    Monday, September 12, 2011 5:47 PM