antenna delay calibration is need or not

[thomasonegd]

Hello guys,in LPS_NODE's TDoA3 positioning system, do all nodes（ including all Anchors and all TAGs）, do not need to perform antenna delay calibration? In the uwb_tdoa_anchor3.c and lpsTdoa3Tag.c files, I see that the antenna delay settings are all zero. Does this assume that the consistency of all DWM1000 modules is better and does not require antenna delay calibration?

If all modules are calibrated for antenna delay, can the positioning accuracy be higher?

Btw

Code: Select all

 double clockCorrection = calculateClockCorrection (anchorCtx, remoteTxSeqNr, remoteTx, rxTime); 

what is the scope of clockCorrection here, and the results of your tests?

[kimberly]

In TDOA3, we assume that all antenna delays are the same and therefore will cancel eachother out (since it is not based on direct ranging but the difference in time of arrive/distance). However, if this assumption does not hold and the delays are very different, this would ofcourse make a difference in quality. But it wasn't really necessary in our experience but would be good to have a solid test to verify this.

There has been some work done on getting a better performance from the LPS system. Here are some exampels:

• https://gitlab.com/jdiep/master-thesis

• https://www.bitcraze.io/2020/04/learnin ... crazyflie/