i bought the MDEK1001 with the PANS release 1 a few weeks ago and have already made some experience with it.
I worked with it in my buro (6x5 m) with 4 anchors and two tags. Now i want to do a test with all 12 modules (10 anchor and 2 tags) on a testarea 20x100 m. I want to position the anchors in a pattern like this: (, ’ , ’ , ’ , ') and the two tags have to move through them. In the future with more anchors the length of the test area should be longer.
I dont just want to capture the tags position with the smartphone app, i wanted to capture it with the tag connected to a pc (with tera term). Now in the new PANS release 2 the Gateways are available. Thats the first question:
Can i put one Gateway on the end or beginning of the testarea and collect the data from the position of the tags, even when the testarea is longer?
Second Question ist about the two way ranging method. With PANS release 1 the double-sided twr method was used, where the twr-slot was 5 ms long. In the new PANS release 2 the twr-slot is just 3 ms long and it uses the single-sided twr method.
Does the ss-twr method has significant higher errors (like mentioned in the dw1000 user manual) than the ds-twr method? Is the uncalibrated dwm1001 PANS2 in range of ± 10-30 cm accuracy? Is it a result of the new IOT-Data slot and the upgrade from 16 to 30 BCN-Slots?
You will need to place enough Gateway to the covered the area. The amount of Gateway depends on its signal coverage. Please keep the modules out of the Fresnel Zones to achieve the best range. It’s out of scope to explain that here in details, but there many articles about this available on the internet.
SS-TWR should provide the same accuracy as DS-TWR. The SS-TWR provides better power consumption, more effective air-time usage allowing to add more BCN slots, IOT data exchange and UWB backhaul for the gateway functionality.
I would also recommend to rather place the anchors in rectangular/square formation to achieve the best position estimation.
So i need for every 20-40 m a Gateway, this means everytime a new Raspi. Sounds really expensive to cover a large area for tracking tags with mqtt. My first understanding of the gateway was that the tags get there positions due to trilateration and send there xyz coordinates+id (iot data) to the anchors, which transmit it to the gateway (iot data). But the gateway is “just” a better listener who have to be in uwb-range of a tag?
Is it possible to send the xyz coords + id over the new iot-data to the anchors an they transmit the data to the gateway on one side of the testarea?
Okay, so if the accuracy should be the same, then i have no problem with the new TWR method.
The formation of the anchors is a result of the given parameters of my test. In between the testarea the tag should be able to see always 3 anchors and sometimes 4. My test should show me the reality if it is close enough or i have to put them really in a rectangular formation.
Yes, the gateway functions as a better listener. Unfortunately it is not possible to send the data to the server via the anchors. Not in the Release 2.
I’m wondering something similar. Is there anything about the gateway that makes it “better” as a listener than a passive node in shell mode? I understand that it adds user data functionality, but is there any advantage from a strictly position data collection POV?
The gateway functionality is designed to work with higher capacity and robustness.
The listener was designed primarily for the demo purposes.
Technically you can use it via UART, but as it has not been designed for high performance, you might hit the limit with increasing number of tags earlier.