Hello,
If I just buy the QM33120W chip, would it be possible to get 360° aoa by using 3 antennas, where they might be in like a triangle shape and the chip will do a reading from antenna 1 and 2, switch from antenna 2 to antenna 3 using some switch, get a reading from that and fuse the data together to solve the front back ambiguity? or is there any better solution?
Thanks in advance,
Bob
Hi Bob,
the UWB beam switching is already a patented solution from Adant Tech, officially a Qorvo partner. Try and check their website at sMART UWB - Adant Technologies Inc., and contact them if interested!
Antenna switching setups can get confusing pretty quickly, especially when dealing with timing and signal loss, so this explanation makes things a lot easier to understand.
Bob,
Sorry for my belated response.
What you describe is basically sound. The concept is that an AOA system which has at least a 120 degree field of view can be made to work over 360 degrees if you can replicate it 3 times to cover the full circle.
There are two ways to replicate it, in space and in time. In space means you use 3 chips so that when a tag sends one packet, all three compute AOA over their wedge of usability. This implies 6 antennas, although you could share each antenna with two chips, but that gets a bit weird on transmit. Or in time, means the tag transmits 3 packets and you use one chip with RF muxes to select which antenna pairs to measure. This can be done with only 3 antennas Either way, you end up with 3 AOA outputs and then you have to resolve where the actual direction is coming from.
There is a bit of math and logic to turn the 3 AOA readings into a single direction since each AOA has a “mirror” image direction as well, and the actual direction might be the mirror one from the back and not the front one you would assume.
Switching around antennas to measure phase of an arriving signal is old prior art, so the basic idea is not patentable in modern times. A specific and particular implementation might be, depends on what the actual patent says.
We have done a lot of work in phase system using UWB. We found the 3 antenna solution to be lacking and gave poor results. We increase this to 5 antennas and that performed MUCH better.
Our (very old!) demo video of the 5 antenna head:
This system uses 5 chips, one for each antenna, and we drive each chip with the same clock to keep them synced. It needs only one packet to determine bearing, and it does two way ranging to determine distance. Since TWR is done with two packets from the tag, we get two bearings and average the results for even better accuracy.
You will notice the antennas are much further apart than half a wavelength (as it typical for most AOA systems). We developed an algorithm that deals with the phase ambiguities and thus has higher angular precision due to the greater antenna spacing. This is the part we patented.
Using 4 or 6 antennas was worse due to the symmetry this causes. 7 antennas also works nicely, but that is getting excessive in antenna and chip count. 3 antennas is just not very good in our experiments. It may work for you if your angular requirements are tolerant of the errors.
We never commercialized this concept but may do so in the future.
Mike Ciholas, President, Ciholas, Inc
3700 Bell Road, Newburgh, IN 47630 USA
mikec@ciholas.com