Athero-CSI-tool-Python-RemoteReceive-Liveview-AmplitudeScaled
Athero-CSI-tool-Python-RemoteReceive-Liveview-AmplitudeScaled copied to clipboard
shuspieler
About scaled csi
Hi! Your tool scaled the amplitude, how about the phase? Does it need to be scaled? Regards.
@WyJ12101 Hi for the phase, Here are some places should be pay attention to.
The fist one is the subcarriers frequency. with the HT40 mode, the bandwidth is 40MHz with 114 subcarriers. The interval is 312.5k and their relationship of offset's index can be found in the IEEE 802.11n-2009 standard (page 50) Notice here that there is a gap between the lower 20Mz and the upper 20Mz. The subcarrier's interval is not equal distance around the canter frequency. If you wish, you can perform some interpolation in the carrier's canter ( there will be 117 subcarriers including the 3 interpolated carriers)
The second notice is the phase jump between the lower 20Mhz and the upper 20Mhz. The upper phase has an +90 degree offset which is a characteristic from the standard. (page 264, 272, 273) (use '90' as keyword to search in the above standard) You can manually remove the offset for a better smooth plot.
There is my cable test (to simulate the ideal environment) for my better explanation.
The third notice is the most troubling one. During my experiment, I observe that when one of the subcarriers has a great fading, thee will be a sudden change in the phase and when the change is greater than 2 Pi, at this subcarrier there will be an extra offset. There is an example.
In the above figure, there are two continuous packets, this subcarrier has a great fading, so between these two packets, the phase has an extra offset at this subcarrier. In my mind this should be an automatic wrapping compensation in the hardware.
There is a clearer example. The transmitter antenna is fixed and the receiver antenna is moving constantly. ID packet on the horizontal axis means the the received CSI packet's sequence, which is also corresponding to the time. The vertical axis means the 114 subcarriers. The different color represent different value of phase. The antenna was moving constantly, thus the phase changing should be smoothly. But as you see on the figure, sometimes there will be the flip of phase and sometime the phase will flip back. This phenomenon occurred frequently if the RF environment is complex. It's a pity I don't have more details about that. You should take care of this problem during your research. If you have any solutions about that, it is very kind of you to share with me.
That's all the important notices about the phase. I hope my comment can help you a bit. It will be very nice of you to give me a 'star'. Thank you.
Thanks very much for replying! @shuspieler
I have met the same question that phase change not smoothly. I am trying to find it out now. I guess maybe some dynamic compensation measures are needed. How do you deal with the raw phase? I now just do a linear fit.
Above picture is one subcarrier after linear fit in the time domain.
@WyJ12101 Hi, sorry for my late replay. I have left he CSI project for about 1 year. I don't have the latest research progress and I also don't have the experiment environment anymore. Thus, I have no advice for your problem. Sorry for that. I hope everything goes well for you in your research.