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.

中文翻译：

---

通过多标签分类进行源检测

无线电源检测问题被重新表述为多类分类问题，并使用深度学习框架解决。使用具有全向元素的中心对称线性阵列对输入波形进行采样，并将自相关矩阵的归一化上三角提取为一维 (1D) CNN 的输入特征，经过训练以在存在的情况下检测源不相关和相关的信号。介绍了检测算法，并随后针对传统的源检测算法进行了基准测试。我们针对具有挑战性的操作条件对算法进行了压力测试，并进行了广泛的评估，以展示所引入的预测模型的有效性和贡献。