The paper "Path signature-based XAI-enabled network time series classification" by Le SUN, Yueyuan WANG, Yongjun REN, and Feng XIA addresses the challenges of classifying network time series (NTS) data, particularly in the context of 5G networks. The authors propose a novel model called Recurrent Signature (RecurSig), which combines deep learning (DL) techniques with explainable artificial intelligence (XAI) to enhance the accuracy and transparency of NTS classification. RecurSig includes two main components: Verbe, a data augmentation module using one-dimensional convolutional neural networks (1D-CNN), and SigRNN, a feature-extraction module equipped with XAI capabilities. The model converts time series data into multidimensional paths, capturing dynamic patterns and generating interpretable features. Extensive experiments on public datasets demonstrate that RecurSig outperforms existing models in both accuracy and explainability, making it a promising solution for network protection and optimization. The paper also reviews related work on XAI technologies and discusses the importance of accurate NTS classification in network automation systems.The paper "Path signature-based XAI-enabled network time series classification" by Le SUN, Yueyuan WANG, Yongjun REN, and Feng XIA addresses the challenges of classifying network time series (NTS) data, particularly in the context of 5G networks. The authors propose a novel model called Recurrent Signature (RecurSig), which combines deep learning (DL) techniques with explainable artificial intelligence (XAI) to enhance the accuracy and transparency of NTS classification. RecurSig includes two main components: Verbe, a data augmentation module using one-dimensional convolutional neural networks (1D-CNN), and SigRNN, a feature-extraction module equipped with XAI capabilities. The model converts time series data into multidimensional paths, capturing dynamic patterns and generating interpretable features. Extensive experiments on public datasets demonstrate that RecurSig outperforms existing models in both accuracy and explainability, making it a promising solution for network protection and optimization. The paper also reviews related work on XAI technologies and discusses the importance of accurate NTS classification in network automation systems.