Its a good effort but very pedestrian and a very low hanging fruit. Its just another academia paper that will be published.
https://doi.org/10.1109/GLOBECOM38437.2019.9014297 https://doi.org/10.1109/CCNC.2018.8319181 https://dl.acm.org/doi/abs/10.1145/3286978.3287003 ..... many more.
I'd say this is far more interesting, does not use ML and credits the tech stacks that it leverages . https://people.csail.mit.edu/davidam/docs/WiMic_final.pdf