I plan to rebuild a positioning algorithm and use a lot of anchors to provide more distance data to improve the positioning accuracy of the tag, but DWM1001 can only use up to four distance data. I want to know how to get more distance data from anchors. I sincerely ask for your help.
Hi @zwj,
PANS RTLS is designed to support only max. of 4 measurements. You would need to create your own system in order to support a higher amount of measurements per cycle.
Depending on your requirements regarding the overall network capacity, Tag power consumption, scalability it can be a relatively simple task (based on the Decawave/Qorvo examples) to a very complex multiple years of development. The question is too generic so it is hard to give you a specific answer.
Cheers,
TDK
1 Like
Hi @zwj , what are your requirements in terms of accuracy ?
I’m trying to achieve the same as you, I want to use TWR and get the distance from a runtime defined number of anchors. I’m providing my code as open source but I’m starting from scratch and still do not have per device antenna calibration or power compensation. If you’re a company and have a budget, then probably the best is to contact one of Decawave partners to do the program for you, but if you’re in academic research or doing this on your own let me know.
@leapslabs .Thank you very much for your reply.
I’m still confused. I have 12 dwm1001 modules, and I try to use 11 as anchors and one as target. I want to get 11 distance data from anchors through target, but I can only get 4. I understand that I may need to modify the dwm1001 firmware to achieve my goal, but I don’t know the source code of the firmware. I can only find a dwm1001 from the official website_ PANS_ R2.0.hex file and dwm1001_ on_ board_ package。 I want to know how to get the source code of dwm1001 and where to modify the source code so that I can get the four data of anchor from target.
Thank you again.
Hi,@wassfila.
Thank you very much for your reply.
I am studying how to use dwm1001 to make the positioning accuracy within 10 mm. I found that using only four can’t achieve my goal, so I want to set anchor several times to try to improve the accuracy, but I don’t know how to modify the firmware of dwm1001.
@zwj , sure here’s the link to my repo
just yesterday, I managed to get any number of TWR sequence running, you just give a list of anchors ids in responders list of the RF packet command sm{"uwb_cmd":"twr","initiator":0,"responders":[1,2,3,5,7],"at_ms":100,"step_ms":10,"count":10,"count_ms":50}
here’s an example with long TWR sequence, each group of pulses are logging a TWR initiator or responder
But from the way you describe the process you’ve been through, I’m not sure if you’re willing to write and maintain your own positioning framework. In case you start from the TWR samples, you have a lot of recommendations to follow (compensation, antenna calibration) before the increase of the number of Anchors can have a valuable impact on your measures.
Did you read this document from the product Documentation section ? “APS011_Sources-of-Error-in-Two-Way-Ranging-Schemes_v1.1.pdf”
In case you decide which source code to you, let me know, I do not recommend to use mine to start with as it’s still a work in progress, unless that’s all you find and you’re ready to improve it. But in case you want to use it I’ll provide help for it.
If you’re aiming for 10mm accuracy on position then you have a lot of work ahead of you. The measured range between two devices will change by several cm if you rotate them on the spot.
I have seen people claim <1cm position errors. But only by using a huge number of anchors, from memory I think they said it was a ~50 anchor TDoA system.
Getting anything under 50 mm position accuracy starts getting tricky.
Hi @zwj
with current PANS version you can get maximally 4 ranges with nearby anchors. The PANS software is only provided as a binary file and the source codes are not public.
Cheers
JK
@wassfila .
Thank you for your sincere help. I wanted to get the source code of PNAs version before, but I got the message that it would not be disclosed.
Thank you again.
1 Like
Hi @zwj
the PANS software is disclosed. It is distributed as a hex image or a linkable library so you can write a small user application on the top of the PANS software. But you can get only 4 ranges here and no more.
Cheers
JK