Mobility performance has been a key focus in cellular networks up to 5G. To enhance handover (HO) performance, the Third Generation Partnership Project introduced a conditional HO and layer 1/layer 2 (L1/L2)-triggered mobility (LTM) mechanisms in 5G. While these reactive HO strategies address the tradeoff between HO failures and ping-pong effects, they often result in inefficient radio resource utilization due to additional HO preparations. To overcome these challenges, this article proposes a proactive HO framework for mobility management in open radio access network (O-RAN), leveraging user-cell link predictions to identify the optimal target cell for HO. We explore various categories of graph neural networks (GNNs) for link prediction and analyze the complexity of applying them to the mobility management domain. Two GNN models are compared using an anonymized real-world dataset and a synthetic dataset obtained with Keysight’s EXata emulation tool. The results show the models’ ability to capture the dynamic and graph-structured nature of cellular networks. Finally, we present key considerations for real-world deployment that outline future steps to enable the integration of GNN-based link prediction for mobility management in O-RAN networks.
Due to the great complexity, heterogeneity, and variety of services, anomaly detection is becoming an increasingly important challenge in the operation of new generations of mobile communications. In many cases, the underlying relationships between the multiplicity of parameters and factors that can cause anomalous behavior are only determined by human expert knowledge. On the other hand, although automatic algorithms have a great capacity to process multiple sources of information, they are not always able to correctly signal such abnormalities. In this sense, this paper proposes the integration of both components in a framework based on Active Learning that enables enhanced performance in anomaly detection tasks. A series of tests have been conducted using an online anomaly detection algorithm comparing the proposed solution with a method based on the algorithm output alone. The obtained results demonstrate that a hybrid anomaly detection model that automates part of the process and includes the knowledge of an expert following the described methodology yields increased performance.
One of the most crucial aspects of cellular networks is coverage, as it determines the areas where users can connect to the network and utilize its services. In the past, the use of planning tools was common practice for the establishment of coverage areas and network capacity prior to the deployment of a network. However, issues with coverage, such as interference or coverage gaps, may arise due to equipment malfunctions, suboptimal configurations, or alterations in the propagation environment. In particular, an inadequate antenna tilt configuration can result in overshoot or undershoot situations on the network, which in turn can give rise to the aforementioned problems. This paper proposes a methodology for the real-time detection of overshoot and undershoot situations. To achieve this goal, KPI (Key Performance Indicators) are analyzed using machine learning techniques. Given the difficulty of detecting coverage problems in mobile networks, the results obtained suggest that the methodology provides a consistent knowledge base for optimizing the antenna tilt, thereby improving network performance.
The arrival of the Fifth Generation (5G) entails a significant evolution in the context of mobile communication networks. This new technology will bring heterogeneous scenarios with new types of services and an increasingly high number of users and nodes. The efficient management of such complex networks has become an important challenge. To address this problem, automatic and efficient algorithms must be developed to facilitate operators’ management and optimization of their networks. These algorithms must be able to cope with a very high number of heterogeneous data and different types of scenarios. In this paper, a novel framework for a cellular network behavioral analysis and monitoring is presented. This framework is based on a combination of unsupervised and supervised machine learning techniques. The proposed system can analyze the behavior of cells and monitor them, searching for behavior changes over time. The information extracted by the framework can be used to improve subsequent management and optimization functions.