Xinchen Lyu

AI-enhanced CSI feedback reduces signaling overhead but introduces reconstruction errors that, together with bursty arrivals, can violate delay guarantees. This letter studies downlink resource allocation under joint traffic uncertainty and AI-reconstructed CSI uncertainty, modeled by Gaussian traffic residuals and log-normal channel perturbations. We derive an ϵ-budgeted tail-bound reformulation that converts the coupled chance constraint into a convex deterministic program. Simulations show the proposed joint design meets ϵ_k = 10⁻³ under stressed uncertainties with moderate power overhead, whereas CSI-only or traffic-only protection fails.

Online model training is pivotal for enabling multiuser semantic communication systems to adapt to dynamic channel conditions. However, conventional frameworks suffer from prohibitive communication overhead and vulnerabilities to privacy attacks, hindering practical deployment. This paper proposes semantic information mixup (SIMix), a secure and efficient training framework that integrates Over-the-Air Mixup (OAM) with label-aware user grouping to jointly optimize spectral efficiency and semantic security. The OAM mixes semantic features of multiple users via wireless channels, inherently obfuscating sensitive data while reducing communication overhead. A closed-form Tx-Rx scaling optimization minimizes the mean square error (MSE) of over-the-air computation under channel noise, ensuring stable convergence in low-SNR regimes. Furthermore, an extended max-clique algorithm dynamically partitions users into groups with minimal intra-label similarity, reducing model inversion attack success rates. Experiments on CIFAR-10 and Tiny ImageNet demonstrate that the proposed approach is superior in terms of communication efficiency and security, reducing communication overhead by up to 25% and attaining 17.58 dB PSNR (20.98 dB reduction) under inversion attack and reducing 13.44% attack success rate under label inference attack, while achieving comparable transmission accuracy.

Wireless foundation models promise transformative capabilities for channel state information (CSI) processing across diverse 6G network applications, yet face fundamental challenges due to the inherent dual heterogeneity of CSI across both scale and scenario dimensions. However, current pretraining approaches either constrain inputs to fixed dimensions or isolate training by scale, limiting the generalization and scalability of wireless foundation models. In this paper, we propose HeterCSI, a channel-adaptive pretraining framework that reconciles training efficiency with robust cross-scenario generalization via a new understanding of gradient dynamics in heterogeneous CSI pretraining. Our key insight reveals that CSI scale heterogeneity primarily causes destructive gradient interference, while scenario diversity actually promotes constructive gradient alignment when properly managed. Specifically, we formulate heterogeneous CSI batch construction as a partitioning optimization problem that minimizes zero-padding overhead while preserving scenario diversity. To solve this, we develop a scale-aware adaptive batching strategy that aligns CSI samples of similar scales, and design a double-masking mechanism to isolate valid signals from padding artifacts. Extensive experiments on 12 datasets demonstrate that HeterCSI establishes a generalized foundation model without scenario-specific finetuning, achieving superior average performance over full-shot baselines. Compared to the state-of-the-art zero-shot benchmark WiFo, it reduces NMSE by 7.19 dB, 4.08 dB, and 5.27 dB for CSI reconstruction, time-domain, and frequency-domain prediction, respectively. The proposed HeterCSI framework also reduces training latency by 53% compared to existing approaches while improving generalization performance by 1.53 dB on average.

Decentralized federated learning (DFL) is pivotal for the integration of AI and communication in 6G, facilitating distributed edge-network learning without centralized coordination. However, existing DFL frameworks suffer from prohibitive communication overhead and slow convergence, with a critical issue being the difficulty of balancing the tradeoffs between communication efficiency and convergence rates. This paper proposes a novel distributed synchronization topology orchestration method for DFL that leverages the diameter of the communication graph as a tunable parameter to optimize this communication-convergence tradeoff. Our key insight stems from the relationship between the topology diameter and the spectral radius of the consensus matrix, i.e., constraining the diameter reduces the spectral radius, thereby accelerating convergence while minimizing communication costs. Specifically, we formulate the synchronization topology design as a diameter-constrained minimum spanning tree (MST) problem, where the diameter serves as an adjustable parameter to balance the tradeoff. To ensure robustness in dynamic networks, we further devise a distributed adaptation mechanism that maintains the diameter-constrained MST despite node/link failures. Experiments on the CIFAR-10 dataset demonstrate that our method achieves accuracy comparable to fully connected topologies while reducing communication costs by 25% and accelerating convergence by 28% against state-of-the-art benchmarks.