About the ranging results of sx1280

[dysun]

Hello,

I am testing 1280 and I find that when I use a single channel to do the ranging 40 times a round in a 50m long distance, sometimes the result will be steady for a whole round, maybe like this:
45.342548
55.686599
56.317608
55.258413
50.1878
55.866887
55.573918
53.90625
55.010517
54.447115
55.010517
52.238582
53.117488
53.432993
51.652644
51.990685
52.080829
50.751202
49.511719
49.060998
49.444111
48.970853
49.151142
49.286358
47.663762
49.060998
48.677885
47.641226
48.001803
46.604567
45.545373
46.041166
44.689002
44.260817
45.951022
45.455228
43.04387
52.734375
43.855168
43.246695

but sometimes the result will be more disturbing, like this:
35.178786
24.947416
17.668269
21.183894
47.663762
23.167067
20.597957
21.364183
50.773738
19.876803
23.617788
20.665565
19.764123
18.209135
19.60637
19.854267
22.423377
17.172476
20.507812
20.890925
19.561298
21.183894
23.099459
22.671274
21.702224
18.366887
18.299279
19.921875
20.214844
18.141526
22.243089
18.907752
19.335938
20.597957
19.651442
15.66256
17.713341
15.978065
20.079627
How could this happen?

Best regards,
Sun Dayang

[tcooper]

Hello Sun,

I suspect that multipath is causing this. Every practical environment will be subject to multipath propagation which, in the majority of cases, will lead to range over estimation. (Although in some configurations underestimation can occur).

Indoor environments and non Line-of-Sight (nLoS) measurements generally experience more severe multipath. I would therefore recommend starting in an outdoor environment to validate your setup (as shown in this app note).

Because there will be outliers in every set of ranging measurements (both over and under estimations). It is also good practice to perform some statistical filtering on the ranging results. This second, more detailed, application note gives details on the filtering employed in the evaluation kit and the diversity techniques you can employ to help ensure a reliable spread of ranging results.

Thanks, Tim

[stuart]

I would concur with the comments about multipath, I have carried out a fair bit of testing of the SX1280 ranging function.

I live in a typical urban area and the reported distances can vary significantly as you move up and down the street outside my house.

Over the nearby local park, which has a open field that is 200m x 300m, reported distances are a lot more stable. As expected really.