Enriched K-Tier Heterogeneous Satellite Networks Model with User Association Policies

In the rapid evolution of the non-terrestrial networks (NTNs), satellite communication has emerged as a focal area of research due to its critical role in enabling seamless global connectivity. In this paper, we investigate two representative user association policies (UAPs) for multi-tier heterogeneous satellite networks (HetSatNets), namely the nearest satellite UAP and the maximum signal-to-interference-plus-noise-ratio (max-SINR) satellite UAP, where each tier is characterized by a distinct constellation configuration and transmission pattern. Employing stochastic geometric, we analyze various intermediate system aspects, including the probability of a typical user accessing each satellite tier, the aggregated interference power, and their corresponding Laplace transforms (LTs) under both UAPs. Subsequently, we derive explicit expressions for coverage probability (CP), non-handover probability (NHP), and time delay outage probability (DOP) of the typical user. Furthermore, we propose a novel weighted metric (WM) that integrates CP, NHP, and DOP to explore their trade-offs in the system design. The robustness of the theoretical framework is verified is verified through Monte Carlo simulations calibrated with the actual Starlink constellation, affirming the precision of our analytical approach. The empirical findings underscore an optimal UAP in various HetSatNet scenarios regarding CP, NHP, and DOP..

Cross-Domain Causal Preference Learning for Out-of-Distribution Recommendation

Recommender systems use users' historical interactions to learn their preferences and deliver personalized recommendations from a vast array of candidate items. Current recommender systems primarily rely on the assumption that the training and testing datasets have identical distributions, which may not hold true in reality. In fact, the distribution shift between training and testing datasets often occurs as a result of the evolution of user attributes, which degrades the performance of the conventional recommender systems because they fail in Out-of-Distribution (OOD) generalization, particularly in situations of data sparsity. This study delves deeply into the challenge of OOD generalization and proposes a novel model called Cross-Domain Causal Preference Learning for Out-of-Distribution Recommendation (CDCOR), which involves employing a domain adversarial network to uncover users' domain-shared preferences and utilizing a causal structure learner to capture causal invariance to deal with the OOD problem. Through extensive experiments on two real-world datasets, we validate the remarkable performance of our model in handling diverse scenarios of data sparsity and out-of-distribution environments. Furthermore, our approach surpasses the benchmark models, showcasing outstanding capabilities in out-of-distribution generalization.