dw1000 range bias

I just measured the range result between two UWB nodes under different conditions.
it’s very accurate under LOS condition, but turns to have a large distance deviation (1~2m) when nodes are blocked by human body or a wall.
My question is : Is it possible to correct this deviation using range bias? the method of estimating the signal power in the first path provided in user manaua seems not accurate enough when RSL is high.

1 Like

Hello, I would like to recommend a paper submitted by me to IPIN 2016,¨Ranging in UWB using commercial radio modules:

experimental validation and NLOS mitigation¨ might be helpful for you.
kind regards

Hi abbas,

thanks for your reply, I have read your paper and list some questions below, any suggestions are appreciated:

  1. NLOS Identification :
    you added two extra conditions:
    If RSL−FSL = 6,7,8,9,10 dB and RSL ≤ −87 dB
    is considered an NLOS channel
    If RSL−FSL = 6,7,8,9,10 dB and RSL > −87 dB
    is considered an LOS channel
    the threshold -87db is based on free space received powers from Friis’ equation.
    Q1: I think RSL is much related with the circumstance(because it’s the total power of all the recieved signals), while FSL is more appropriate to compare with free space received power.
    Q2: There seems to no way to get the exact FSL/RSL when they are above -85dBm using dw1000, see Figure 22 “Estimated RX level versus actual RX level” in dw1000 user manual.

  2. NLOS Mitigation:
    the formula: d = d’ − w(Ri − 1)
    Q3: How can we get the w and Ri when the target are moving?

[color=#333333]thanks for your reply.[/color]

[color=#333333]Q1: I think RSL is much related with the circumstance(because it’s the total power of all the received signals), while FSL is more appropriate to compare with free space received power.[/color]

Answer: Yes, sure [color=#333333]FSL is more appropriate to compare [/color]with free[color=#333333] space received power. In EVK 1000 kit, you are not able to obtain free space received powers unless you reprogram it. So, if EVK1000 as it, you have to deal with these parameters FSL and RSL. The experiment is evaluated by specificity and sensitivity measurements and the accuracy is increased as you can see in the paper.[/color]

[color=#333333]Q2: There seems to no way to get the exact FSL/RSL when they are above -85dBm using dw1000, see Figure 22 “Estimated RX level versus actual RX level” in dw1000 user manual.[/color]
I could not understand what do you mean by FSL/RSL do you mean FSL and RSL or division. Anyway, please look at DW user manual 2015.
[color=#333333]2. NLOS Mitigation:[/color]
[color=#333333]the formula: d = d’ − w(Ri − 1) [/color]
[color=#333333]Q3: How can we get the w and Ri when the target are moving?[/color]
[color=#333333]Answer[/color]
[color=#333333]We already used a map for that. The experiment was not for positioning. It is to validate the DW in such environment. [/color]

I just mean FSL and RSL, the way to evaluate FSL and RSL provided in DW user manual is not accurate, [color=#333333]see Figure 22 “Estimated RX level versus actual RX level” in dw1000 user manual.[/color]