Search

Autonomous network management is crucial for Fifth Generation (5G) and Beyond 5G (B5G) networks, where a constantly changing environment is expected and network configuration must adapt accordingly. Modeling tools are required to predict the impact on performance (packet and delay loss) when new traffic demands arrives and when changes in routing paths are applied in the network. Mathematical analysis and network simulators are techniques for modeling networks but both have limitations, as the former provides low accuracy and the latter requires high execution times. To overcome these limitations, machine learning (ML) algorithms, and more specifically, graph neural networks (GNNs), are proposed for network modeling due to their ability to capture complex relationships from graph-like data while predicting network properties with high accuracy and low computational requirements. However, one of the main issues when using GNNs is their lack of generalization capability to larger networks, i.e., when trained in small networks (in number of nodes, paths length, links capacity), the accuracy of predictions on larger networks is poor. This paper addresses the GNN generalization problem by the use of fundamental networking concepts. Our solution is built from a baseline GNN model called RouteNet (developed by Barcelona Neural Networking Center-Universitat Politècnica de Catalunya (BNN-UPC)) that predicts the average delay in network paths, and through a number of simple additions significantly improves the prediction accuracy in larger networks. The improvement ratio compared to the baseline model is 101, from a 187.28% to a 1.828%, measured by the Mean Average Percentage Error (MAPE). In addition, we propose a closed-loop control context where the resulting GNN model could be potentially used in different use cases. [ABSTRACT FROM AUTHOR]

The increasing proportion of bifacial photovoltaic modules (Bi-PVM) in new projects makes the operation of photovoltaic system (PVS) more complicated, and it is difficult to accurately predict the power of the PVS. To solve this problem, this paper proposes a new power prediction method for PVS based on Bi-PVM. Firstly, the equal proportion digital twin model of the example project is constructed. The superposition principle is used to analyze the factors affecting the power generation performance of Bi-PVM, and the characteristic project is constructed according to the analysis results. Secondly, the bifacial correction coefficient is introduced to reduce the parameter error caused by Bi-PVM to the prediction model. On this basis, a power prediction machine learning model based on bidirectional gated recurrent unit (Bi-GRU) network is established. Finally, a simulation experiment is carried out on the TensorFlow machine learning platform. With the actual operation data of a PV power station in Jiuquan, China, the simulation analysis is carried out under four weather types, namely, sunny day, rainy day, snowy day and complex and changeable day, respectively, which verified the correctness and excellence of the proposed method. [ABSTRACT FROM AUTHOR]