You are not logged in.
Hi,
I was trying out the channel estimate viewer with log data obtained from two WARP boards.
I was looking for a case where magnitude varies greatly among subcarriers, but can't seem to obtain that kind of log.
Even when I place the receiver far away or behind walls, the magnitude is almost flat in the graphs.
The maximum difference I saw was around 5dB. It is also the case with the example log, and also the figure shown in the webpage.
I am wondering if I am missing something or looking at wrong stats. Some papers, such as ones using CSI for localization fingerprinting, show frequency response graphs where the magnitude significantly vary among subcarriers (15-20dB).
https://arxiv.org/pdf/1603.07080.pdf
Has anyone seen these kind of lines with WARP?
Thanks.
Offline
A few ideas:
-The logs will only contain RX_OFDM entries for packets which had valid preambles and PHY headers. The PHY can't decode packets with invalid preamble/headers, so the MAC has no way to log these events. Channels with large amplitude variations across frequency are harder to decode, so there will be fewer RX_OFDM log entries in high-frequency-selective channels.
-Channels with more multipath will be more frequency-selective. Moving your client to another room will increase multipath effects, but it might still result in a channel with a dominant path. You could also try adding reflectors to create additional paths of comparable amplitude.
Offline