ABSTRACT Introduction With the development and integration of satellite and terrestrial networks, mobile traffic prediction has become more important than before, which is the basis for service provision and resource scheduling when supporting various vertical applications. However, existing traffic prediction methods, especially deep learning‐based methods, require massive data for model training. Due to data privacy concerns, mobile traffic data are not easily shared among different parties, making it difficult to obtain a precise prediction model. Methods To mitigate the data leakage risk, a federated learning framework is proposed in this study for mobile traffic prediction in satellite‐terrestrial integrated networks to achieve a tradeoff between data privacy and prediction accuracy. In the proposed framework, local models are trained in base stations on the ground, and a global model is aggregated in the satellite edge server in space. Results A deep learning‐based prediction model with an adaptive graph convolutional network (AGCN) and long short‐term memory (LSTM) modules is proposed and validated in numerical experiments, which achieves the lowest prediction error with a real‐world traffic dataset when compared with other graph neural network (GNN) variants in the federated learning setting. Conclusion Numerical experiments with a real‐world mobile traffic dataset demonstrate the effectiveness of the proposed approach, which outperforms other GNN variants with lower prediction errors.

Satellite-terrestrial integrated networks (STINs) have been recognized as an important key technology for future 6G networks. With the introduction of large-scale satellite constellations, traditional network management schemes cannot fully meet the requirements for network operation and software defined networking (SDN) is introduced as a promising tool. However, existing network simulators fail to meet the simulation purposes for software-defined STINs and we aim to fill this research gap by implementing a network simulation framework with open-source simulation platforms including Geomview, SaVi and Mininet.

With the increase in satellite meg-constellations, network management has become increasingly complex, and the software-defined network idea has been introduced into modern satellite networks. This study considers the multi-controller placement problem in software-defined satellite networks and decomposes the problem into two sub-problems: SDN controller placement and switch-controller assignment. A meta-heuristic approach based on AVOA is proposed to solve the two sub-problems step-by-step, with the optimization objective of network reliability maximization, which has often been neglected in previous studies. Numerical experiments demonstrate that the AVOA-based solution outperforms baselines based on the NSGA-II and particle swarm optimization algorithms in terms of control delay, load-balancing capability, and reliability.

Internet of Things (IoT) is an important infrastructure for supporting vertical applications. However, existing IoT systems are still facing some challenges, e.g., lack of coverage in rural areas and lack of efficient data collection methods. To overcome these challenges, a satellite IoT framework is proposed in this study as a promising solution, and smart agriculture is used as a typical application scenario. A satellite edge computing workflow is further proposed, with computer vision as a case study. In the case study, a lightweight deep learning model named MobileViT is proven effective for aphid detection and infestation severity classification on lemon leaves.

Drought has been a global concern and an effective prediction method is needed. Meteorological data are seen as an efficient and economic approach. Challenges arise with the large volume and high nonlinearity between meteorological variables and the drought level. In this study, deep learning is proposed as an effective solution for drought level prediction as multivariate time series classification. The synthetic minority oversampling technique is further adopted to alleviate the class imbalance problem and improve the classification performance. Experimental results on an open dataset named DroughtED demonstrate the effectiveness of the proposed deep learning method.