The proliferation of low-Earth orbit (LEO) satellite constellations presents unprecedented opportunities for distributed machine learning (ML) applications. However, the inherent challenges of sparse connectivity, heterogeneous communication windows, and non-independent and identically distributed (non-IID) data across satellites hinder the effectiveness of conventional federated learning (FL) frameworks. To address these challenges, we propose Model Contrastive Federated Learning (MCFL), a novel framework tailored for LEO satellite constellations. MCFL introduces a two-stage approach: 1) similarity-based satellite clustering to mitigate intra-cluster data imbalance by grouping satellites with aligned data distributions, and 2) collaborative staleness-aware learning that employs semi-asynchronous model aggregation within clusters to balance convergence speed and model accuracy. The key contributions include a contrastive loss function for robust representation learning under class imbalance, gradient sparsification to minimize communication overhead, and an inter-cluster knowledge-sharing mechanism to prevent cluster-specific model bias. Extensive simulations on the EuroSAT dataset demonstrate that MCFL achieves an improvement of 15% in test accuracy and reduces training time $3\times $ compared to state-of-the-art FL baselines while reducing communication costs by 40%. This work bridges the gap between distributed learning theory and practical satellite constraints, offering a scalable solution for real-time ML applications in dynamic space-terrestrial networks.
Ren Ozeki, Mohamed Rihan, Hamada Rizk et al.· IEEE Open Journal of the Com...· 0 citations
This paper proposes a novel physical-layer security framework for multi-UAV Integrated Sensing and Communication (ISAC) networks operating in adversarial environments. To maximize the sum secrecy rate of legitimate ground users (GUs) while satisfying minimum sensing beampattern-gain constraints for target illumination, we introduce a dynamic role allocation mechanism in which each UAV can switch, on a per-time-slot basis, between an ISAC mode—combining coherent communications with radar sensing—and a dedicated artificial noise (AN) jammer mode. The resulting optimization is cast as a highly coupled Mixed-Integer Non-Linear Program (MINLP) that jointly optimizes binary role indicators, transmit beamforming and sensing covariance matrices, and UAV trajectories. We solve this problem with a tailored Alternating Optimization (AO) algorithm that integrates a penalty-based Convex-Concave Procedure (CCP) for the binary role subproblem, Semidefinite Relaxation (SDR) for the beamforming subproblem, and a trust-region Successive Convex Approximation (SCA) for the trajectory subproblem. Numerical results demonstrate that the proposed dynamic-role architecture consistently outperforms both a static dedicated-jammer scheme and a fully optimized all-ISAC embedded-AN benchmark, confirming that its secrecy advantage arises from adaptive spatial-functional specialization rather than from artificial-noise transmission alone. Furthermore, we characterize the fundamental tradeoff between secrecy performance and stringent sensing beampattern-gain constraints, showing that moderate sensing requirements can be accommodated with no secrecy penalty.
M. M. Selim, Mohamed Rihan, Armin Dekorsy et al.· IEEE Open Journal of the Com...· 0 citations
