With the significant increase of communication network scales, intelligent Network Operation and Maintenance (NOM) becomes essential. Knowledge Graphs (KGs) are a key enabler for intelligent NOM, and Relational Triple Extraction (RTE) plays a critical role in KG construction. However, most existing RTE researches rely on general-domain corpora, with limited exploration into the specialized domain. In this paper, we identify a novel challenge in Chinese NOM corpus —Segmented Entity, which has garnered little attention in prior works. To address it, this paper proposes an Entity-level Autoregressive RTE (EARTE) method, which incorporates an innovative Segmented-BIO (Begin, Inside, Outside) tagging scheme. Furthermore, we construct the CMIM23-NOM1-RA, the first high-quality restricted domain RTE dataset for NOM. Throughout the experimentation, we meticulously reproduce all baselines and provide a comprehensive analysis. The results show that EARTE achieves the best performance on CMIM23-NOM1-RA. EARTE’s F1 scores surpass those of the best-performing baselines by 0.4%, 2.7%, and 0.8% under the strict criterion, the lenient criterion, and the setting focusing only on segmented entities, respectively. Finally, our codes, dataset, and reproduction guidelines are publicly available at: https://github.com/JYzzzzzz/PEAR-RTE.

Network Operation and Maintenance (NOM) is an important part of the IT and telecommunications infrastructure. To improve the efficiency of NOM personnel, predict base station alarms, and classify base stations more precisely, we build a new labeled time series dataset, called BSAlarm, which collects 1000 different time series from processed real-world network alarm log data. Furthermore, we develop a NOM pipeline that employs B-spline interpolation alongside the moving average method to create and utilize this dataset for both long- and short-time series forecasting (LTSF), as well as base station classification tasks. Unlike prior NOM methods, our pipeline is centered on the generation of time series data, thereby enhancing alarm prediction performance and the visualization of changes under base station alarm conditions. Extensive experiments show that the proposed BSAlarm dataset and pipeline achieve remarkable results on time-series-data-based NOM.

Inter-cell interference poses a significant challenge to the performance and reliability of cellular networks due to the complex spatial and temporal relationships between network nodes. Addressing this issue requires accurate prediction and assessment of interference. This paper presents a novel solution leveraging the strengths of a weighted graph convolutional network (WGCN) combined with graph coloring techniques. Specifically, we propose a WGCN-based interference estimation model to accurately derive the real-time inter-cell interference. Then, a graph multi-coloring problem is considered for the interference coordination. To address the color collision between cells and the color (i.e. spectrum resources) requirement of individual cells in the graph coloring problem, we propose a WGCN-assisted graph multi-coloring (WGCN-GMC) algorithm to allocate spectrum resources rationally. Simulation results demonstrate that our approach significantly enhances interference coordination, and achieves an impressive average improvement of 58.2 % compared to the traditional GMC algorithm leading to improved overall network performance.

By integrating machine learning with edge devices, we can augment the capabilities of edge devices, such as IoT devices, household appliances, and wearable technologies. These edge devices generally operate on microcontrollers with inherently limited resources, such as constrained RAM capacity and limited computational power. Nonetheless, they often process data in a high-velocity stream fashion, exemplified by sequences of activities and statuses monitored by advanced industrial sensors. In practical scenarios, models must be interpretable to facilitate troubleshooting and behavior understanding. Implementing machine learning models on edge devices is valuable and challenging, striking a balance between model efficacy and resource constraint. To address this challenge, we introduce our novel Onfesk, which combines online learning algorithms with an innovative interpretable kernel. Specifically, our Onfesk trains an online classifier over the kernel's feature sketches. Benefiting from our specially designed modules, the kernel's feature sketches can be efficiently produced, and the memory requirements of the classifier can be significantly reduced. As a result, Onfesk delivers effective and efficient performance in environments with limited resources without compromising on model interpretability. Extensive experiments with diverse real-world datasets have shown that Onfesk outperforms state-of-the-art methods, achieving up to a 7.4% improvement in accuracy within identical memory constraints.

Root cause analysis for faults is one of the core tasks in the operation and maintenance of communication networks. Although artificial intelligence techniques can be used to assist manual inspections, fault diagnosis is still a tough problem. In the scenario of fault root cause localization, insufficient associated information makes it difficult to accurately determine the root cause. Meanwhile, it is a big challenge to extract as many feature details as possible from limited information and then fully utilize them. Therefore, this paper proposes a Knowledge-Enhanced Transformer-FL method, namely, KETrans-FL, to address the problem of root cause localization, by treating it as a multi-class classification problem. Our method first constructs a knowledge graph for knowledge enhancement, which consists of four types of nodes (base station, alarm, fault and alarm level) and their relationships based on the network operations. This knowledge enhancement technique incorporates real operation data and other external knowledge (e.g., alarm level) to serve as the source of feature information and thus can extract statistical and embedded features. Then, our method designs a Trans-FL (Transformer-Focal Loss) model, which uses the adapted Transformer encoder to learn the correlation information between input features to generate classification probabilities and employs Focal Loss as the loss function to mitigate severe class imbalance in the multi-class classification problem. Experimental results show that our proposed KETrans-FL method achieves a classification accuracy of nearly 91% and an average AUC score of 93%, indicating a significant improvement on fault root cause localization compared to baseline models. In addition, experimental results also validate the remarkable effect of our knowledge enhancement technique on improving the final classification accuracy.

Computing set difference cardinalities is a critical task in database optimization, network management, and anomaly detection. Due to the limited computational and mem-ory resources, exactly calculating set difference cardinalities becomes impractical in real-world applications. To solve this issue, sketch methods such as Odd sketch, Tug-of-War sketch, and HyperLogLog sketch can be extended to provide approximate estimations of set difference cardinalities. They use a family of hash functions to compress all elements in a set into a compact data structure. Unfortunately, Odd sketch suffers from limited estimation range, while Tug-of-War sketch and HyperLogLog sketch unavoidably face the problems of large estimation errors and high computational costs. In this paper, we design a novel data structure of bit array GXBits to fast and accurately estimate set difference cardinalities in a large range. In GXBits, the prob-ability of each bit recording its corresponding elements follows a variant of geometric distributions and varies across different bits. We conduct extensive experiments on synthetic datasets and real-world datasets. Experimental results demonstrate that our method GXBits is more computationally and memory efficient, and significantly increases the estimation accuracy of existing methods by up to 221.3 times.

Graphs are widely used to represent the relations among entities. When one owns the complete data, an entire graph can be easily built, therefore performing analysis on the graph is straightforward. However, in many scenarios, it is impractical to centralize the data due to data privacy concerns. An organization or party only keeps a part of the whole graph data, i.e., graph data is isolated from different parties. Recently, Federated Learning (FL) has been proposed to solve the data isolation issue, mainly for Euclidean data. It is still a challenge to apply FL on graph data because graphs contain topological information which is notorious for its non-IID nature and is hard to partition. In this work, we propose a novel FL framework for graph data, FedCog, to efficiently handle coupled graphs that are a kind of distributed graph data, but widely exist in a variety of real-world applications such as mobile carriers’ communication networks and banks’ transaction networks. We theoretically prove the correctness and security of FedCog. Experimental results demonstrate that our method FedCog significantly outperforms traditional FL methods on graphs. Remarkably, our FedCog improves the accuracy of node classification tasks by up to 14.7%.