We present a unified framework for semantic communication under bounded privacy leakage, in which an encoder must reveal selected information about a source while restricting the disclosure of correlated private data. In contrast to prior work that merely injects noise over a fixed representation, we co-design both the semantic mapping and the noise mechanism. Leveraging extended versions of the Functional Representation Lemma (FRL) and the Strong Functional Representation Lemma (SFRL), we model how the disclosed data arise from the original source and correlated private data. We then formulate a new optimization problem to align the resulting distributions with a 'goal' distributionbalancing accuracy for the user's task and privacy constraints on sensitive data. Furthermore, we propose two distinct mapping strategies that map the original signal domain to a compact semantic space. Our numerical results verify the effectiveness of this joint design, demonstrating significant benefits over conventional methods for privacy-constrained semantic communication.

With numerous renewable generators and energy storage systems integrated into the power grids, the security-constrained DC optimal power flow (DCOPF) is essential for power system operation. For large-scale power grids, traditional CPU-based optimization algorithms (such as the simplex and barrier methods) have saturated in computational efficiency and are inherently difficult to parallelize. To tackle these issues, by incorporating the symmetric Gauss–Seidel (sGS) decomposition, this work develops a GPU-based Halpern Peaceman-Rachford algorithm, termed the sGS-HPR, which enjoys an O(1k) iteration complexity in terms of the KKT residual. Moreover, the closed-form solutions for all subproblems are derived, which only consist of matrix- vector multiplications and vector operations, and thus can be easily parallelized on GPUs. As a consequence, the developed sGS-HPR algorithm enjoys a O(N_L × n/ϵ) non-ergodic computational complexity in terms of floating-point operations for obtaining an ϵ-optimal solution measured by the KKT residual for large-scale DCOPF problems, where n represents the variable dimension, and N_L denotes the number of branches in the power grid. Extensive numerical tests on large-scale power grids, reaching up to the 9241- bus PEGASE system, demonstrate the scalability and superior efficiency of the developed GPU-based sGS-HPR algorithm compared to state-of-the-art methods. Notably, the proposed method achieves a 6× speedup compared with Gurobi for large-scale instances. Additionally, for ultra-large-scale cases, Gurobi throws an “out-of-memory” error, while the proposed sGS-HPR algorithm maintains its computational scalability and efficiency.

We study the performance of the Thompson Sampling algorithm for logistic bandit problems. In this setting, an agent receives binary rewards with probabilities determined by a logistic function, $\exp(\beta \langle a, \theta \rangle)/(1+\exp(\beta \langle a, \theta \rangle))$, with parameter $\beta>0$, and where both the action $a\in \mathcal{A}$ and the unknown parameter $\theta \in \mathcal{O}$ lie within the $d$-dimensional unit ball. Adopting the information-theoretic framework introduced by Russo and Van Roy, we derive regret bounds via the analysis of the information ratio, a statistic that quantifies the trade-off between the immediate regret incurred by the agent and the information it gained about the parameter $\theta$.We improve upon previous results and establish that the information ratio is bounded in $O\big(d \alpha^{-2} \big)$, where $d$ is the dimension of the problem and $\alpha$ is a \emph{minimax measure} of the alignment between the action space $\mathcal{A}$ and the parameter space $\mathcal{O}$. Notably our bound does not scale exponentially with the logistic slope and is independent of the cardinality of the action and parameter spaces. Using this result, we derive a bound on the Thompson Sampling expected regret of order $O(d \alpha^{-1} \sqrt{T \log(\beta T/d)})$, where $T$ is the number of time steps. To our knowledge, this is the \emph{first regret bound for any logistic bandit algorithm} that eliminates any exponential scaling with $\beta$ and is independent of the number of actions. In particular, when the parameters are on the sphere and the action space contains the parameter space, the expected regret bound is of order $O(d \sqrt{T \log(\beta T/d)})$.

This paper studies the Bayesian regret of the Thompson Sampling algorithm for bandit problems, building on the information-theoretic framework introduced by Russo and Van Roy [1]. Specifically, it extends the rate-distortion analysis of Dong and Van Roy [2], which provides near-optimal bounds for linear bandits. A key limitation of these results is the assumption of a finite action space. We address this by extending the analysis to settings with infinite and continuous action spaces. Additionally, we specialize our results to bandit problems with expected rewards that are Lipschitz continuous with respect to the action space, deriving a regret bound that explicitly accounts for the complexity of the action space.

This paper studies the Bayesian regret of a variant of the Thompson-Sampling algorithm for bandit problems. It builds upon the information-theoretic framework of [1] and, more specifically, on the rate-distortion analysis from [2], where they proved a bound with regret rate of O(d T log(T))for the d-dimensional linear bandit setting. We focus on bandit problems with a metric action space and, using a chaining argument, we establish new bounds that depend on the metric entropy of the action space for a variant of Thompson-Sampling. Under suitable continuity assumption of the rewards, our bound offers a tight rate of O(d√T) for d-dimensional linear bandit problems.

In this paper, we investigate the performance of an integrated sensing and communication (ISAC) system within a cell-free massive multiple-input multiple-output (MIMO) system. Each access point (AP) operates in the millimeter-wave (mmWave) frequency band. The APs jointly serve the user equipments (UEs) in the downlink while simultaneously detecting a target through dedicated sensing beams directed toward a reconfigurable intelligent surface (RIS). Although the AP-RIS, RIS-target, and AP-target channels have both line-of-sight (LoS) and non-line-of-sight (NLoS) parts, only knowledge of the LoS paths is assumed to be available. A key contribution of this study is the consideration of clutter, which degrades the target detection performance if not handled. We propose an algorithm to alternatively optimize the transmit power allocation and the RIS phase-shift matrix, maximizing the target signal-to-clutter-plus-noise ratio (SCNR) while ensuring a minimum signal-to-interference-plus-noise ratio (SINR) for the UEs. Numerical results demonstrate that exploiting the clutter subspace significantly enhances detection probability, particularly at high clutter-to-noise ratios, and reveal that an increased number of transmit side clusters impairs detection performance. Finally, we highlight the performance gains achieved using a dedicated sensing beam.

To enhance straggler resilience in federated learning (FL) systems, a semi-decentralized approach has been recently proposed, enabling collaboration between clients. Unlike the existing semi-decentralized schemes, which adaptively adjust the collaboration weight according to the network topology, this letter proposes a deterministic coded network that leverages wireless diversity for semi-decentralized FL without requiring prior information about the entire network. Furthermore, the theoretical analyses of the outage and the convergence rate of the proposed scheme are provided. Finally, the superiority of our proposed method over benchmark methods is demonstrated through comprehensive simulations.

In this paper, we investigate the problem of distributed learning (DL) in the presence of Byzantine attacks. For this problem, various robust bounded aggregation (RBA) rules have been proposed at the central server to mitigate the impact of Byzantine attacks. However, current DL methods apply RBA rules for the local gradients from the honest devices and the disruptive information from Byzantine devices, and the learning performance degrades significantly when the local gradients of different devices vary considerably from each other. To overcome this limitation, we propose a new DL method to cope with Byzantine attacks based on coded robust aggregation (CRA-DL). Before training begins, the training data are allocated to the devices redundantly. During training, in each iteration, the honest devices transmit coded gradients to the server computed from the allocated training data, and the server then aggregates the information received from both honest and Byzantine devices using RBA rules. In this way, the global gradient can be approximately recovered at the server to update the global model. Compared with current DL methods applying RBA rules, the improvement of CRA-DL is attributed to the fact that the coded gradients sent by the honest devices are closer to each other. This closeness enhances the robustness of the aggregation against Byzantine attacks, since Byzantine messages tend to be significantly different from those of honest devices in this case. We theoretically analyze the convergence performance of CRA-DL. Finally, we present numerical results to verify the superiority of the proposed method over existing baselines, showing its enhanced learning performance under Byzantine attacks.

In this paper, we consider the problem of federated learning (FL) with devices that have intermittent connectivity to the central server. For this problem, the concept of semi-decentralized FL has been proposed in the literature. This paradigm allows non-straggler devices to relay the gradients computed by the stragglers to the server, and enables realization of gradient coding (GC) to mitigate the negative impact of the stragglers that fail to communicate directly to the central server. However, for GC in semi-decentralized FL, the communication overhead caused by information transmission among the devices is significant. To overcome this shortcoming, inspired by the existing communication-optimal exact consensus algorithm (CECA), we propose a new communication-efficient semi-decentralized FL method (COFFEE). In each round, the devices exchange information by taking a certain number of steps towards communication-optimal exact consensus, ensuring that each device obtains the average of the gradients computed by both its previous neighbors and itself. Afterwards, the non-stragglers transmit the local average result to the server for global aggregation to update the global model. We analyze the convergence performance and the communication overhead of COFFEE analytically. Building on this, to further enhance learning performance under a specific communication overhead, we propose an enhanced version of COFFEE with an adaptive aggregation rule at the central server, referred to as A-COFFEE, which adjusts to the straggler pattern of the devices over training rounds. Experiments are conducted to verify that the proposed methods outperform the baseline methods.

This work studies gradient coding (GC) in the context of distributed training problems with unreliable communication. We propose cooperative GC (CoGC), a novel gradient-sharing-based GC framework that leverages cooperative communication among clients. This approach eliminates the need for dataset replication, making it communication- and computation-efficient and suitable for federated learning (FL). By employing the standard GC decoding mechanism, CoGC yields strictly binary outcomes: the global model is either recovered exactly or the recovery is meaningless, with no intermediate outcomes. This characteristic ensures the optimality of the training and demonstrates strong resilience to client-to-server communication failures. However, due to the limited flexibility of the recovery outcomes, the decoding mechanism may also result in communication inefficiency and hinder convergence, especially when communication channels among clients are in poor condition. To overcome this limitation and further exploit the potential of GC matrices, we propose a complementary decoding mechanism, termed GC⁺, which leverages information that would otherwise be discarded during GC decoding failures. This approach significantly improves system reliability against unreliable communication, as the full recovery¹ of the global model dominates in GC⁺. To conclude, this work establishes solid theoretical frameworks for both CoGC and GC⁺. We assess the system reliability by outage analyses and convergence analyses for each decoding mechanism, along with a rigorous investigation of how outages affect the structure and performance of GC matrices. Finally, the effectiveness of CoGC and GC⁺ is validated through extensive simulations.