Jumping height of lec output of tag despite of constant tag height

Wireless Connectivity Ultra-Wideband
1

Hello,
I used several anchors that are configured with a different height. Moreover, I have one listener that is connected to the computer and one active tag that is moving around and has almost a constant height (~1.70). I wonder why in my output of lec my tag has varying heights that are much to high (e.g. 2.10 or 1.95).

POS,0,C704,-0.51,6.36,2.10,97,x06
POS,0,C704,-0.39,6.31,2.10,99,x06
POS,0,C704,-0.17,6.28,1.95,98,x06
POS,0,C704,-0.00,6.29,2.07,96,x06
POS,0,C704,0.24,6.28,1.96,95,x06
POS,0,C704,0.27,6.29,2.19,93,x06

Could it be possible that you cannot have anchors which have not the same height? Do I have to many anchors (there are 7 anchors and one of them is 5 meters higher than the others that only vary in ~1m)?
Thanks in advance!
Best regards,
Blubbus

2

Hi blubbus,

The tag will range with four anchors when calculating the position. You may want to make sure that it is actually selecting the anchor which is 5 meters high.

The others anchors are separated by about a meter. With that kind of geometric dilution, we can expect an error on the Z axis.

It seems that you are measuring a 30cm error offset against the estimate of 1.7m, and then measurement are within 15cm.

The 15cm is a good metric for this system. I am not completely sure why there is such an offset. If your objective is mainly to measure position along Z axis, I would recommend to maybe place anchors lower, maybe 50cm and see if it improves the offset.

Make sure the tag is always in los condition with the anchors.

Let us know how it goes,

Thanks
Yves

3

Hey,
thank you for the fast reply. I have to add that the error is much more than 30cm since it varies in 4 meters. All the anchors are places with a big distance to each other but the height of those anchors differs from 7.5m to 1.12m.

    A1(0,0,2.15)  --- A2(0,13,1.30) --- A3(0, 27, 2,15)
  /

A0(9.5, - 6, 7.5)
\
A4(23, 0, 2.12) — A5(23, 13, 2,12) — A6(23,27, 1.12)

Is something wrong with that setup? We are using the lates Firmware v9 on those chips.
Thank you so much for replying!
Best regards,
Blubbus

4

Do you think that this is a conceptual problem using GDOP like described in the last comment of that post?

5

Hi bubbus,

Yes, it is very possible your network is subject to this.

A0 is fairly high, but all the other anchors are within 1.5 meters of each of them (along the Z axis).

From your logs it look like all the Z coordinates are within 1.95m and 2.20m, thus the 30cm. Where is you 4m error figure coming from ?

Thanks
Yves

6

Hey,
my log file is much longer and there are some values that have a low z-achsis:
1555081810301
,POS,0,C704,11.96,6.83,0.88,97,x06
1555081810402
,POS,0,C704,12.15,6.92,0.74,95,x06
1555081810502
,POS,0,C704,12.28,6.83,0.73,97,x06
1555081810602
,POS,0,C704,12.44,6.81,0.66,96,x06
1555081810702
,POS,0,C704,nan,nan,nan,0,x06
1555081810802
,POS,0,C704,12.65,6.82,0.67,96,x06

and

1555081850802
,POS,0,C704,4.86,15.95,0.63,94,x07
1555081850902
,POS,0,C704,4.81,15.82,0.48,94,x07
1555081851002
,POS,0,C704,4.77,15.69,0.40,77,x07
1555081851102
,POS,0,C704,4.71,15.58,0.26,91,x07
1555081851202
,POS,0,C704,4.64,15.44,0.12,89,x07
1555081851302
,POS,0,C704,4.66,15.25,0.83,85,x07
1555081851402

And some that have a high z-value:
1555081857802
,POS,0,C704,3.49,4.91,3.38,98,x07
1555081857902
,POS,0,C704,3.62,4.72,3.11,93,x07
1555081858002
,POS,0,C704,3.47,4.59,3.37,97,x07
1555081858102
,POS,0,C704,3.29,4.46,3.70,95,x07
1555081858202
,POS,0,C704,3.44,4.28,3.55,96,x07
1555081858302
,POS,0,C704,3.53,4.13,3.47,95,x07

Do you think reordering the anchors in a tetrahedron or a cube with highly different heights might reduce the huge error? Is this high error normal? It seems that the z-value increases if you come closer to A0.
Thank you so much for helping me!
Best regards,
Blubbus