Large-scale low Earth orbit (LEO) remote satellite constellations have become a brand new, massive source of space data. Federated learning (FL) is considered a promising distributed machine learning technology that can communicate optimally using these data. However, when applying FL in satellite networks, it is necessary to consider the unique challenges brought by satellite networks, which include satellite communication, computational ability, and the interaction relationship between clients and servers. This study focuses on the siting of parameter servers (PSs), whether terrestrial or extraterrestrial, and explores the challenges of implementing a satellite federated learning (SFL) algorithm equipped with client selection (CS). We proposed an index called “client affinity” to measure the contribution of the client to the global model, and a CS algorithm was designed in this way. A series of experiments have indicated the advantage of our SFL paradigm—that satellites function as the PS—and the availability of our CS algorithm. Our method can halve the convergence time of both FedSat and FedSpace, and improve the precision of the models by up to 80%.

With the rapid evolution of artificial intelligence (AI) models, remote sensing applications increasingly rely on computationally intensive deep learning algorithms. However, the outdated onboard computing hardware and limited processing capacity of a single satellite can no longer meet these growing demands. While offloading data to the ground for processing is a common alternative, it faces significant challenges due to bandwidth-limited and unstable satellite-to-ground communication links. To overcome these limitations, we propose a flexible framework that leverages existing in-orbit satellite computing resources to accommodate AI models of varying scales and complexities. As a representative application scenario, this paper focuses on in-orbit object detection of remote sensing images. By aggregating idle computing resources from multiple satellites, the proposed framework enables distributed onboard image processing, which significantly reduces the volume of data that needs to be transmitted to the ground. To achieve this, we partition convolutional neural networks (CNN) models into multiple sequential slices and deploy them across satellites, taking into account model complexity, hardware heterogeneity, and dynamic inter-satellite links (ISLs). We formulate the model partitioning problem as a Series Multi-Armed Bandit (SMAB) problem and propose both online and offline algorithms to solve it. Compared to the traditional method of transmitting raw remote sensing data to ground stations via ISLs for post-processing, our approach reduces end-to-end latency by 35%.

Although most aircraft are equipped with active communication systems such as ADS-B and satellite links, these systems may fail in emergencies or when signals are lost. At this point, remote sensing, as a passive observation method that does not require a target response, can provide crucial visual information and effectively gaps left by traditional methods. With the support of existing remote sensing technologies, aircraft targets in remote sensing imagery can now be effectively identified and extracted through using image features. However, when reconstructing sub-diffraction-limited targets (e.g., aircraft registration numbers), key high-frequency information is irreversibly lost during imaging due to the optical system's inherent diffraction limit. This significantly restricts image interpretability. To address this limitation, we propose a new super-resolution reconstruction framework that combines physical optical modeling with deep diffusion models. Firstly, a parallel collimated light simulation system is constructed to model the degradation process due to diffraction and generate training samples that characterize the multi-level blurring effect. Secondly, the collected multi-frame training samples with varying degrees of blurriness are used as the conditional input to the diffusion model to enhance feature representation. Inspired by the principle of wavefront modulation and its compensation, we have designed two dedicated modules: the Selective Channel Attention Unit (SCAU) and the Spatial Gated Feature Enhancement (SGFE), which explicitly compensate for the high-frequency attenuation caused by diffraction. The experimental results demonstrate that this method outperforms existing methods on both public remote sensing datasets and self-collected real-world datasets. It is effective at reconstructing identity information, such as aircraft registration numbers, from aerial remote sensing images, expanding the range of identities detectable by current remote sensing satellites.