As 6G networks proliferate, they generate vast volumes of data and engage diverse devices, pushing the boundaries of traditional network management techniques. The limitations of these techniques underpin the need for a revolutionary shift towards AI/ML-based frameworks. This paper introduces a transformative approach using our novel Speed-optimized LSTM (SP-LSTM) model, an embodiment of this crucial paradigm shift. We present a proactive strategy integrating predictive analytics and dynamic routing, underpinning efficient resource utilization and optimal network performance. This innovative, two-tiered system combines SP-LSTM networks and Reinforcement Learning (RL) for forecasting and dynamic routing. SP-LSTM models, boasting superior speed, predict potential network congestion, enabling preemptive action, while RL capitalizes on these forecasts to optimize routing and uphold network performance. This cutting-edge framework, driven by continuous learning and adaptation, mirrors the evolving nature of 6G networks, meeting the stringent requirements for ultra-low latency, ultra-reliability, and heterogeneity management. The expedited training and prediction times of SP-LSTM are game-changers, particularly in dynamic network environments where time is of the essence. Our work marks a significant stride towards integrating AI/ML in future network management, highlighting AI/ML's exceptional capacity to outperform conventional algorithms and drive innovative performance in 6G network management.

Rapid and precise identification of Coronavirus Disease 2019 (COVID-19) is pivotal for effective patient care, comprehending the pandemic’s trajectory, and enhancing long-term patient survival rates. Despite numerous recent endeavors in medical imaging, many convolutional neural network-based models grapple with the expressiveness problem and overfitting, and the training process of these models is always resource-intensive. This paper presents an innovative approach employing Xception, augmented with cutting-edge transfer learning techniques to forecast COVID-19 from X-ray thorax images. Our experimental findings demonstrate that the proposed model surpasses the predictive accuracy of established models in the domain, including Xception, VGG-16, and ResNet. This research marks a significant stride toward enhancing COVID-19 detection through a sophisticated and high-performing imaging model.

Seismic full wave inversion (FWI) is a widely used non-linear seismic imaging method used to reconstruct subsurface velocity images, however it is time consuming, has high computational cost and depend heavily on human interaction. Recently, deep learning has accelerated it’s use in several data-driven techniques, however most deep learning techniques suffer from overfitting and stability issues. In this work, we propose an edge computing-based data-driven inversion technique based on supervised deep convolutional neural network to accurately reconstruct the subsurface velocities. Deep learning based data-driven technique depends mostly on bulk data training. In this work, we train our deep convolutional neural network (DCN) (UNet and InversionNet) on the raw seismic data and their corresponding velocity models during the training phase to learn the non-linear mapping between the seismic data and velocity models. The trained network is then used to estimate the velocity models from new input seismic data during the prediction phase. The prediction phase is performed on a resource-constrained edge device such as Raspberry Pi. Raspberry Pi provides real-time and on-device computational power to execute the inference process. In addition, we demonstrate robustness of our models to perform inversion in the presence on noise by performing both noise-aware and no-noise training and feeding the resulting trained models with noise at different signal-to-noise (SNR) ratio values. We make great efforts to achieve very feasible inference times on the Raspberry Pi for both models. Specifically, the inference times per prediction for UNet and InversionNet models on Raspberry Pi were 22 and 4 s respectively whilst inference times for both models on the GPU were 2 and 18 s which are very comparable. Finally, we have designed a user-friendly interactive graphical user interface (GUI) to automate the model execution and inversion process on the Raspberry Pi.

An unrestricted, global, and ubiquitous connectivity, available to anyone, everywhere, and anytime, is a key objective of the 6G vision. A true, seamless integration of terrestrial (T) and non-terrestrial (NT) networks is mandatory to reach such an ambitious target. From this perspective, the three-year ITA-NTN project (2023-2025), funded by the European Union within the framework of NextGenerationEU work program, aims to investigate advanced solutions for the seamless integration of Terrestrial and Non-Terrestrial networks. This paper considers the project aspects concerning the provision of seamless connectivity across the heterogeneous T and NT layers by proposing an integrated ecosystem, which enables transparent information exchange among the various communicating nodes belonging to each layer. The ecosystem relies on the software-based management of fully reconfigurable network architectures based on disruptive concepts like cell-free multiple-input multiple-output (MIMO), full software-defined radio (SDR)-based multi-waveforms physical layer design, use of aggressive capacity-oriented radio resource management strategies based on Non Orthogonal Multiple Access (NOMA), extended use of Software-Defined Networking (SDN), and Artificial Intelligence (AI) to monitor and manage the entire network architecture. The discussion of the achieved results validates the proposed ecosystemic approach both in terms of viability and quantitative performance improvement.

We study Unmanned Aerial Vehicle (UAV) assisted 5G uplink connectivity for disaster response, where a UAV acts as an aerial base station to restore service to ground users. We formulate a joint control problem coupling UAV kinematics (bounded acceleration and velocity), per-subchannel uplink power allocation, and uplink Non-Orthogonal Multiple Access (UL-NOMA) scheduling with adaptive successive interference cancellation (SIC) under a minimum user-rate constraint. The wireless channel follows 3GPP urban macro (UMa) with probabilistic Line of Sight/Non Line of Sight (LoS/NLoS), realistic receiver noise and noise figure, and user equipment (UE) transmit-power limits. We propose a bounded-action proximal policy optimization with generalized advantage estimation (PPO-GAE) agent that parameterizes acceleration and power with squashed distributions and enforces feasibility by design. Across four user distributions (clustered, uniform, ring, edge-heavy) and multiple rate thresholds, our method increases the fraction of users meeting the target rate by 8.2−10.1 percentage points over strong baselines (OFDMA with heuristic placement, PSO-based placement/power, PPO without NOMA) while reducing median UE transmit power by 64.6%. Results are averaged over ≥5 random seeds with 95% confidence intervals; ablations isolate the gains from NOMA, adaptive SIC order, and bounded action parameterization. We discuss robustness to imperfect SIC and CSI errors, and release code/configurations to support reproducibility.

Additive manufacturing (AM) monitoring is fundamentally constrained by the severe scarcity of annotated data for layer-wise segmentation. This paper addresses this bottleneck by introducing a scalable, high-fidelity synthetic data generation pipeline built on the Slice-100K dataset, capable of producing large volumes of layer-wise semantic segmentation masks. Through analysis of this large-scale synthetic data, we identify a systemic foreground–background class imbalance (1:24 ratio) inherent to AM monitoring, which causes standard Dice loss formulations to diverge catastrophically into a phenomenon we formalize as the “Dice Crash.” To effectively leverage large amounts of unlabeled data, we adapt the Meta Pseudo-Labeling (MPL) framework for industrial segmentation. We evaluate MPL’s true marginal utility by integrating it with both a standard U-Net and a robust state-of-the-art nnU-Net architecture. Experimental outputs show that while MPL yields substantial performance gains (+15.2%) on weak baselines, integrating it with an optimally configured strong baseline consistently improves segmentation accuracy and suppresses false foreground detections, thereby mitigating confirmation bias. These findings demonstrate that semi-supervised learning via continuous bilevel optimization offers a practical and robust enhancement to data-scarce additive manufacturing monitoring. Because any hidden defects in the topmost layer will be permanently buried by subsequent extrusion, this foundational layer-wise segmentation step is the most critical primitive of the monitoring pipeline.