Reduce channel sounding set up time.

Hi, I am working on a channel sounding positioning system using one initiator and 7 reflectors. Since concurrent channel sounding does not work, the program performs sequential channel sounding, after one device has finished ranging it sends the result to my laptop through Bluetooth GATT, then it disconnects and the initiator moves on to the next one. I am following the setup protocol from the sample, the channel sounding with ras, and it takes about 5 seconds at least to start ranging, which makes the whole system impractical if the target were to start moving. Is there any way I can cut down the setup time for the channel sounding procedure or is the sample the only way? Another question is that when I am testing with 7 reflectors, I have noticed that I almost never get distance readings from all of them, whereas when I was just testing with 3 I could consistently get readings. Since I am doing sequential readings and disconnecting afterwards, the readings should not be impacted at all since the initiator is still holding the same amount of connections(2, one with laptop to send message and one with the reflector). The only reason I can think of is that I now have 4 more devices advertising in the background while the channel sounding is performing, would these devices advertising in the background cause such a large effect that they don't return data? When I am testing with reflectors they are all facing the initiator at different angles, would that also worsen the effect? But I am only testing them in 8m radius circle. Could it be that since I have a buffer for every single device in the program, it does not have enough RAM to receive stable data? Any feedback is appreciated, thanks.

Parents
  • Hi, 

    The issues you are experiencing are known limitations of the current channel sounding samples and stack support. See this post

    Regards,
    Amanda H.

  • Hi Amanda, I believe the difference is that I am trying to conduct sequential channel sounding rather than concurrent. I have got the system working for 3 reflectors, however, it is when I raise that number to 7 when it stops working. This is the question I am trying to wrap my head around. Because if I am doing sequential anyway, the number of reflectors should not affect the performance since the initiator is still only performing channel sounding with one reflector at a time. My suspicions are either there is now too much interference in the background, or that since I am storing a buffer for each individual reflector on the initiator, all the RAM is being used and thus leaving no space for the channel sounding procedure. Thank you so much for your time.

    Sincerely,

    Triscuit

Reply
  • Hi Amanda, I believe the difference is that I am trying to conduct sequential channel sounding rather than concurrent. I have got the system working for 3 reflectors, however, it is when I raise that number to 7 when it stops working. This is the question I am trying to wrap my head around. Because if I am doing sequential anyway, the number of reflectors should not affect the performance since the initiator is still only performing channel sounding with one reflector at a time. My suspicions are either there is now too much interference in the background, or that since I am storing a buffer for each individual reflector on the initiator, all the RAM is being used and thus leaving no space for the channel sounding procedure. Thank you so much for your time.

    Sincerely,

    Triscuit

Children
Related