Re: Figuring out different multipath components

[Federico 'Larroca' La Rocca]

Hi, 

As Jeff says, it all depends on the sampling rate. If the time difference between paths is smaller than the sampling period, then there's no hope of recovering each gain individually. In your case, since you have wi-fi, you have a somewhat large sampling rate, but in indoor environments the difference in time of arrival is typically smaller. 

Note furthermore that going from inverting the CSI of wi-fi to actually recovering the paths' gain, even if they are resolved by the sampling resolution, is not straightforward. You may check Gallager's great book for a very clear explanation (free at https://ocw.mit.edu/courses/6-450-principles-of-digital-communications-i-fall-2006/resources/book_9/, particularly section 9.4).

best

Federico

El vie, 8 mar 2024 a las 11:53, Jeff Long (<willcode4@gmail.com>) escribió:

The received signal is a combination of a large number of multipath signals, which can not be separated out. It's sort of like a hash of the environment - a one way calculation. Training signals can be used to determine the multipath response and minimize or take advantage of it, but the individual paths are lost.

On Fri, Mar 8, 2024 at 9:36 AM Sourya Saha <saha.sourya17@gmail.com> wrote:

Hi,

This is my first post to this list. I am quite new to GNU Radio. I have a wifi transmitter and a receiver in a room. The signals from the WiFi transmitter get relfected through the various walls and then reach the receiver, except the LOS signal. What i want to do is to calculate the AoA and RSSI of each of the signals from each of the multipaths. 

Could someone suggest how I can do that? Help is really appreciated.