Semantic Knowledge Distillation for Onboard Satellite Earth Observation Image Classification

This study presents an innovative dynamic weighting knowledge distillation (KD) framework tailored for efficient Earth observation (EO) image classification (IC) in resource-constrained settings. Utilizing EfficientViT and MobileViT as teacher models, this framework enables lightweight student models, particularly ResNet8 and ResNet16, to surpass 90% in accuracy, precision, and recall, adhering to the stringent confidence thresholds necessary for reliable classification tasks. Unlike conventional KD methods that rely on static weight distribution, our adaptive weighting mechanism responds to each teacher model's confidence, allowing student models to prioritize more credible sources of knowledge dynamically. Remarkably, ResNet8 delivers substantial efficiency gains, achieving a 97.5% reduction in parameters, a 96.7% decrease in FLOPs, an 86.2% cut in power consumption, and a 63.5% increase in inference speed over MobileViT. This significant optimization of complexity and resource demands establishes ResNet8 as an optimal candidate for EO tasks, combining robust performance with feasibility in deployment. The confidence-based, adaptable KD approach underscores the potential of dynamic distillation strategies to yield high-performing, resource-efficient models tailored for satellite-based EO applications. The reproducible code is accessible on our GitHub repository.

On-board Satellite Image Classification for Earth Observation: A Comparative Study of Pre-Trained Vision Transformer Models

Remote sensing image classification is a critical component of Earth observation (EO) systems, traditionally dominated by convolutional neural networks (CNNs) and other deep learning techniques. However, the advent of Transformer-based architectures and large-scale pre-trained models has significantly shifted, offering enhanced performance and efficiency. This study focuses on identifying the most effective pre-trained model for land use classification in onboard satellite processing, emphasizing achieving high accuracy, computational efficiency, and robustness against noisy data conditions commonly encountered during satellite-based inference. Through extensive experimentation, we compared traditional CNN-based models, ResNet-based models, and various pre-trained vision Transformer models. Our findings demonstrate that pre-trained Transformer models, particularly MobileViTV2 and EfficientViT-M2, outperform models trained from scratch in accuracy and efficiency. These models achieve high performance with reduced computational requirements and exhibit greater resilience during inference under noisy conditions. While MobileViTV2 excelled on clean validation data, EfficientViT-M2 proved more robust when handling noise, making it the most suitable model for onboard satellite Earth observation tasks. In conclusion, EfficientViT-M2 is the optimal choice for reliable and efficient remote sensing image classification in satellite operations, achieving 98.76\% accuracy, precision, and recall. Specifically, EfficientViT-M2 delivered the highest performance across all metrics, excelled in training efficiency (1,000s) and inference time (10s), and demonstrated greater robustness (overall robustness score at 0.79).

Seamless 5G Automotive Connectivity with Integrated Satellite Terrestrial Networks in C-Band

This paper examines integrated satellite-terrestrial networks (ISTNs) in urban environments, where terrestrial networks (TNs) and non-terrestrial networks (NTNs) share the same frequency band in the C-band which is considered the promising band for both systems. The dynamic issues in ISTNs, arising from the movement of low Earth orbit satellites (LEOSats) and the mobility of users (UEs), are addressed. The goal is to maximize the sum rate by optimizing link selection for UEs over time. To tackle this challenge, an efficient iterative algorithm is developed. Simulations using a realistic 3D map provide valuable insights into the impact of urban environments on ISTNs and also demonstrates the effectiveness of the proposed algorithm.

The Impact of LoRA Adapters for LLMs on Clinical NLP Classification Under Data Limitations

Fine-tuning Large Language Models (LLMs) for clinical Natural Language Processing (NLP) poses significant challenges due to the domain gap and limited data availability. This study investigates the effectiveness of various adapter techniques, equivalent to Low-Rank Adaptation (LoRA), for fine-tuning LLMs in a resource-constrained hospital environment. We experimented with four structures-Adapter, Lightweight, TinyAttention, and Gated Residual Network (GRN)-as final layers for clinical notes classification. We fine-tuned biomedical pre-trained models, including CamemBERT-bio, AliBERT, and DrBERT, alongside two Transformer-based models. Our extensive experimental results indicate that i) employing adapter structures does not yield significant improvements in fine-tuning biomedical pre-trained LLMs, and ii) simpler Transformer-based models, trained from scratch, perform better under resource constraints. Among the adapter structures, GRN demonstrated superior performance with accuracy, precision, recall, and an F1 score of 0.88. Moreover, the total training time for LLMs exceeded 1000 hours, compared to under 6 hours for simpler transformer-based models, highlighting that LLMs are more suitable for environments with extensive computational resources and larger datasets. Consequently, this study demonstrates that simpler Transformer-based models can be effectively trained from scratch, providing a viable solution for clinical NLP tasks in low-resource environments with limited data availability. By identifying the GRN as the most effective adapter structure, we offer a practical approach to enhance clinical note classification without requiring extensive computational resources.

An Experimental Study of C-Band Channel Model in Integrated LEO Satellite and Terrestrial Systems

This paper studies the channel model for the integrated satellite-terrestrial networks operating at C-band under deployment in dense urban and rural areas. Particularly, the interference channel from the low-earth-orbit (LEO) satellite to the dense urban area is analyzed carefully under the impact of the environment's characteristics, i.e., the building density, building height, and the elevation angle. Subsequently, the experimental results show the strong relationships between these characteristics and the channel gain loss. Especially, the functions of channel gain loss are obtained by utilizing the model-fitting approach that can be used as the basis for studying future works of integration of satellite and terrestrial networks (ISTNs).

User-Centric Beam Selection and Precoding Design for Coordinated Multiple-Satellite Systems

This paper introduces a joint optimization framework for user-centric beam selection and linear precoding (LP) design in a coordinated multiple-satellite (CoMSat) system, employing a Digital-Fourier-Transform-based (DFT) beamforming (BF) technique. Regarding serving users at their target SINRs and minimizing the total transmit power, the scheme aims to efficiently determine satellites for users to associate with and activate the best cluster of beams together with optimizing LP for every satellite-to-user transmission. These technical objectives are first framed as a complex mixed-integer programming (MIP) challenge. To tackle this, we reformulate it into a joint cluster association and LP design problem. Then, by theoretically analyzing the duality relationship between downlink and uplink transmissions, we develop an efficient iterative method to identify the optimal solution. Additionally, a simpler duality approach for rapid beam selection and LP design is presented for comparison purposes. Simulation results underscore the effectiveness of our proposed schemes across various settings.

A Hybrid Optimization and Deep RL Approach for Resource Allocation in Semi-GF NOMA Networks

Semi-grant-free non-orthogonal multiple access (semi-GF NOMA) has emerged as a promising technology for the fifth-generation new radio (5G-NR) networks supporting the coexistence of a large number of random connections with various quality of service requirements. However, implementing a semi-GF NOMA mechanism in 5G-NR networks with heterogeneous services has raised several resource management problems relating to unpredictable interference caused by the GF access strategy. To cope with this challenge, the paper develops a novel hybrid optimization and multi-agent deep (HOMAD) reinforcement learning-based resource allocation design to maximize the energy efficiency (EE) of semi-GF NOMA 5G-NR systems. In this design, a multi-agent deep Q network (MADQN) approach is employed to conduct the subchannel assignment (SA) among users. While optimization-based methods are utilized to optimize the transmission power for every SA setting. In addition, a full MADQN scheme conducting both SA and power allocation is also considered for comparison purposes. Simulation results show that the HOMAD approach outperforms other benchmarks significantly in terms of the convergence time and average EE.

Harnessing the Power of Swarm Satellite Networks with Wideband Distributed Beamforming

The space communications industry is challenged to develop a technology that can deliver broadband services to user terminals equipped with miniature antennas, such as handheld devices. One potential solution to establish links with ground users is the deployment of massive antennas in one single spacecraft. However, this is not cost-effective. Aligning with recent \emph{NewSpace} activities directed toward miniaturization, mass production, and a significant reduction in spacecraft launch costs, an alternative could be distributed beamforming from multiple satellites. In this context, we propose a distributed beamforming modeling technique for wideband signals. We also consider the statistical behavior of the relative geometry of the swarm nodes. The paper assesses the proposed technique via computer simulations, providing interesting results on the beamforming gains in terms of power and the security of the communication against potential eavesdroppers at non-intended pointing angles. This approach paves the way for further exploration of wideband distributed beamforming from satellite swarms in several future communication applications.

Energy-Efficient Precoding and Feeder-Link-Beam Matching Design for Bent-Pipe SATCOM Systems

This paper proposes a joint optimization framework for energy-efficient precoding and feeder-link-beam matching design in a multi-gateway multi-beam bent-pipe satellite communication system. The proposed scheme jointly optimizes the precoding vectors at the gateways and amplifying-and-matching mechanism at the satellite to maximize the system weighted energy efficiency under the transmit power budget constraint. The technical designs are formulated into a non-convex sparsity problem consisting of a fractional-form objective function and sparsity-related constraints. To address these challenges, two iterative efficient designs are proposed by utilizing the concepts of Dinkelbach's method and the compress-sensing approach. The simulation results demonstrate the effectiveness of the proposed scheme compared to another benchmark method.

Integrated Access and Backhaul via Satellites

To allow flexible and cost-efficient network densification and deployment, the integrated access and backhaul (IAB) was recently standardized by the third generation partnership project (3GPP) as part of the fifth-generation new radio (5G-NR) networks. However, the current standardization only defines the IAB for the terrestrial domain, while non-terrestrial networks (NTNs) are yet to be considered for such standardization efforts. In this work, we motivate the use of IAB in NTNs, and we discuss the compatibility issues between the 3GPP specifications on IAB in 5G-NR and the satellite radio regulations. In addition, we identify the required adaptation from the 3GPP and/or satellite operators for realizing an NTN-enabled IAB operation. A case study is provided for a low earth orbit (LEO) satellite-enabled in-band IAB operation with orthogonal and non-orthogonal bandwidth allocation between access and backhauling, and under both time- and frequency-division duplex (TDD/FDD) transmission modes. Numerical results demonstrate the feasibility of IAB through satellites, and illustrate the superiority of FDD over TDD transmission. It is also shown that in the absence of precoding, non-orthogonal bandwidth allocation between the access and the backhaul can largely degrades the network throughput.