LPS: Positioning of the LPS-Nodes for best performance

[Che.]

Hi,

I am currently experimenting with the LPS for getting best performance out of it.
The Dillution of Precision is my reference point. However I have some questions:
What part of the system can I interpret as the transmitter, and what is the Receiver, i.e. to calculate the actual value of DOP (see Formular with x,y,z = position of receiver, and xi,yi,zi = position of sattelite [GPS])?

Do you have already results of your experiments, where to best place the nodes?
My humble experiences (I did not test it very extensive) a rough error of about 30 to 50 centimeters in all 3 Dimensions, I placed the Nodes on the floor with about 20 centimeters distance from the ground.

This is my experimental setup:


I plan to attach the Nodes to the ceiling.

Thanks for any hints!

Regards, Che.

[whoenig]

Hi Che,

In the current setup, both Crazyflie and anchors are transmitting and receiving, so I don't think you can apply the formula directly.

As for the placement:
* Try asymmetric setups, otherwise your localization algorithm (particle filter or LMS) might get stuck in a wrong local minima.
* You will get better vertical precision if some of the anchors are higher. You can use tripods/microphone stands/speaker stands for that purpose. Arnaud had good results by using 3 near the ceiling and 3 near the ground.

I think the error in all 3 dimensions will be about +/- 20cm with a good setup. For better results, you would need to fuse some other data (e.g. IMU or barometer).

Best,

Wolfgang

[Che.]

Hi Wolfgang,

thanks for your quick answer. I will heed your tips in my new setup and let you know how it works.

Best,
Timo