Page 1 of 1

number of anchors needed

Posted: Tue Mar 05, 2019 3:32 pm
by wydmynd
hi
I found reference to an article in the blog where researchers used LPS system in area of approx 7x14m divided into 4 rooms, using only 6 anchors.
https://www.bitcraze.io/2019/02/using-t ... d-mapping/

i'm planning to buy the LPS but could not find reference to answer -
- how can I know the number of anchors for such an area (and layout) that can still achieve +-10cm accuracy most of the time? at lease in xy plane
- is there a maximum number of anchors? i saw reference to 8 anchors as recommended maximum.

reference articles i found with Decawave chip - 5 anchors over large area. 90% of values have under 25cm error.
https://ieeexplore.ieee.org/abstract/document/7377440
image.png
thanks

Re: number of anchors needed

Posted: Thu Mar 07, 2019 3:36 pm
by tobias
There are many ifs and buts when it comes to the number of anchors. If you run TWR mode like in the article you are referring two, 6 anchors will get you pretty far but then you are limited to one Crazyflie (or possibly more if you run TWR-TDMA). If you run TDoA (basically unlimited amount of crazyflies) we recommend at least 8 anchors, but now accuracy is more sensitive in regard to how you place your anchors.

Re: number of anchors needed

Posted: Tue Mar 19, 2019 12:01 pm
by wydmynd
thank you for the reply tobias. I researched the topic but missed the part about accuracy. Now I read it again. I would love if you had some more details , even for a test case. the documentation refers to a room about 3*6 meters in size, 2.6 meters height. what accuracy can I expect in this case with 6, 8 anchors? what does it mean that tdoa mode is more sensitive to anchor positioning than twr mode? how accurate do I need to know the position of the anchor to get reasonable results?

anyway I will buy one set with 8 anchors and try to report back with accuracy data...
regrads