In some NLOS circumstances first path of the received signal may get buried in noise and this will cause LDE algorithm mark wrong point as first path. I’m looking for a way to ensure the RX timestamp reported by LDE algorithm is true or wrong. What are the ways that you use to get over this?
There are two RX times reported, RX_STAMP and RX_RAWST.
I’ve implemented checks that these two must be within a certain amount of each other. If the difference is over a threshold I assume the leading edge detection messed up and throw the measurement away.
This cleared up the occasional errors I was seeing but I wasn’t in a particularly harsh environment.