My research team and I tried to do this and in the end decided to implement it in sound waves which is much slower and allows to have a longer time delay. At the speed at which a wifi travels it would be hard to detect that the signal arrived ns first to a sensor that to the other. However with sound the difference is big enough to have different times of arrival. The product is here https://www.copsonic.com/, they advertise it as Accurate Indoor GPS.

Dear Park,

what kind of signal are you using for sensing? I'm working on an indoor positioning system that exploits an OFDM symbol combined to the Zadoff-Chu sequences. From the phase analysis of the received signals, I'm able to extract the direction of the movement.

Please feel free to contact me if you wish further details or would like to read my publications on this topic.

Try infrared. It works

Renaud Di Francesco I should make the system using only Wi-Fi sensing. Infrared things are not included in my works.

Victor Mendez Thanks to your reply. Ah, if I want to distinguish the four directions on the X and Y axis, should I set up at least three scanners? (not only 2 scanner as I draw?)@

@Park Juhyeon to detect movement in a plane, mathematically you need at least 3 sensors. However we used 4. Assume the Z as the average distance from the floor the moving object has

Not only can you distinguish, but given enough resources you can do WiFi imaging - see our recent paper for more information

Preprint ISAC Imaging by Channel State Information using Ray Tracing ...