Noise Cancellation using Linrad and the Phase Coherent Capabilities of an RSPduo
The RSPduo is the latest product from SDRplay, and it's main defining feature is it's dual RX channels. The two channels allow you to simultaneously listen to two stations that may be at opposite ends of the spectrum. Of course the same could be done cheaper with two separate RSP1A devices, however the advantage of the RSPduo is that the two channels are phase coherent. This opens up a lot of technical possibilities such as active noise cancelling/spatial filtering of an interfering signal or unwanted RF noise.
The basic idea behind phase coherent noise cancelling is that you have one antenna pointed towards your signal of interest, and the other pointing towards the interfering signal or coupled to the unwanted noise source. The unwanted signal/noise can then be subtracted from the signal of interest, resulting in a cleaner signal. This can only be done if the two channels are phase coherent like they are on the RSPduo.
Over on YouTube and his blog, ICAS Enterprises has been experimenting with noise cancelling on his RSPduo [Part 1] [Part 2]. The blog is entirely in Japanese, but all the relevant screenshots are in English, and we have provided Google Translated links. SDRplay have also provided a description of the process on their blog. The idea is to use Linrad to process the sound output of the two RSPduo channels from SDRuno. Linrad has capabilities that allow you to subtract two signals from one another.
In his experiments he receives a known 50 MHz CW beacon, and then injects an interfering test signal into the system. Then using Linrads 'signal adaption' feature he is able to completely remove the interference.
Why not record the interference (repetitive man-made coherent noise e.g. 50/60Hz switching noise) and play it back subtracting it from the signal during a qso using software algorithm and adjusting phase/timing manually or automatically to remove 50/60hz components???
The concept works, and it’s not new. See http://www.mfjenterprises.com/Product.php?productid=MFJ-1026
The only new thing about it is that we’re using software to do what we used to use hardware to accomplish.
really?
wow
There seems to have been a lot of discussion about phase coherence on the LimeSDR recently; the phase alignment code was merged into the main branch of LimeSuite in 18.04.
No idea if that actually works in practice. Just maybe worth playing around with.
A similar thing can be done with phase cancelling antennas (yagi and ground plane) joined with a phasing harness. Cool that we can do that inside the receiver these days.