Optimizing sparse RFI prediction using deep learning
Optimizing sparse RFI prediction using deep learning
复制标题
DOI：
10.1093/mnras/stz1865
复制
发表时间：
2019-02
影响因子：
4.8
通讯作者：
Joshua Kerrigan;P. Plante;S. Kohn;J. Pober;J. Aguirre;Zara Abdurashidova;P. Alexander;Z. Ali
中科院分区：
文献类型：
--
作者：
Joshua Kerrigan;P. Plante;S. Kohn;J. Pober;J. Aguirre;Zara Abdurashidova;P. Alexander;Z. Ali
射频干扰（RFI）是射电望远镜中始终存在的限制因素，即使在最偏远的观测地点。当希望保留最大量的灵敏度并减少再电离时代研究的污染时，RFI的识别和去除尤为重要。除了改进RFI识别，我们还必须考虑RFI识别算法的计算效率，因为无线电干涉仪阵列（如氢时代再电离阵列（HERA））的接收器数量越来越大。为了解决这个问题，我们提出了一个深度全卷积神经网络（DFCN），它在干涉数据的使用方面是全面的，其中幅度和相位信息都被联合用于识别RFI。我们使用包含模拟RFI的模拟HERA vibration来训练网络，产生一个已知的“地面实况”数据集，用于评估各种RFI算法的准确性。DFCN模型的评估是在67个培养皿HERA-67的观测结果上进行的，并实现了每GPU每小时1.6 × 105 HERA时间排序的1024个通道病毒的数据吞吐量。我们确定，相对于仅振幅网络，包括可见性相位增加了重要的相邻时频上下文，这增加了RFI和非RFI之间的区分。当应用于我们的HERA-67观察时，在预测时包含相位实现了0.81的召回率，0.58的精确度和0.75的F2得分。
Radio frequency interference (RFI) is an ever-present limiting factor among radio telescopes even in the most remote observing locations. When looking to retain the maximum amount of sensitivity and reduce contamination for Epoch of Reionization studies, the identification and removal of RFI is especially important. In addition to improved RFI identification, we must also take into account computational efficiency of the RFI-Identification algorithm as radio interferometer arrays such as the Hydrogen Epoch of Reionization Array (HERA) grow larger in number of receivers. To address this, we present a deep fully convolutional neural network (DFCN) that is comprehensive in its use of interferometric data, where both amplitude and phase information are used jointly for identifying RFI. We train the network using simulated HERA visibilities containing mock RFI, yielding a known ‘ground truth’ data set for evaluating the accuracy of various RFI algorithms. Evaluation of the DFCN model is performed on observations from the 67 dish build-out, HERA-67, and achieves a data throughput of 1.6 × 105 HERA time-ordered 1024 channelled visibilities per hour per GPU. We determine that relative to an amplitude only network including visibility phase adds important adjacent time–frequency context which increases discrimination between RFI and non-RFI. The inclusion of phase when predicting achieves a recall of 0.81, precision of 0.58, and F2 score of 0.75 as applied to our HERA-67 observations.