The problem of radio source detection is reformulated as a multi-class classification problem and solved using deep learning frameworks. Incoming waveforms are sampled using a centro-symmetric linear array with omni-directional elements and the normalized upper triangle of the autocorrelation matrix is extracted as the input feature to an uni-dimensional (1D) CNN, trained to detect the sources in the presence of both uncorrelated and correlated signals. The detection algorithms are introduced and subsequently benchmarked against the conventional source detection algorithms. We stress test the algorithms for challenging operational conditions and present extensive evaluations to show the efficacy and contributions of the introduced predictive models.

Source detection via multi-label classification