Chengcheng Zhao

Jianping He, Lin Cai, Chengcheng Zhao et al.

Privacy-preserving average consensus aims to guarantee the privacy of initial states and asymptotic consensus on the exact average of the initial value. In existing work, it is achieved by adding and subtracting variance decaying and zero-sum random noises to the consensus process. However, there is lack of theoretical analysis to quantify the degree of the privacy protection. In this paper, we introduce the maximum disclosure probability that the other nodes can infer one node's initial state within a given small interval to quantify the privacy. We develop a novel privacy definition, named $(ε, δ)$-data-privacy, to depict the relationship between maximum disclosure probability and estimation accuracy. Then, we prove that the general privacy-preserving average consensus (GPAC) provides $(ε, δ)$-data-privacy, and provide the closed-form expression of the relationship between $ε$ and $δ$. Meanwhile, it is shown that the added noise with uniform distribution is optimal in terms of achieving the highest $(ε, δ)$-data-privacy. We also prove that when all information used in the consensus process is available, the privacy will be compromised. Finally, an optimal privacy-preserving average consensus (OPAC) algorithm is proposed to achieve the highest $(ε, δ)$-data-privacy and avoid the privacy compromission. Simulations are conducted to verify the results.

Bing Liu, Boao Kong, Limin Lu et al.

Decentralized learning often involves a weighted global loss with heterogeneous node weights $λ$. We revisit two natural strategies for incorporating these weights: (i) embedding them into the local losses to retain a uniform weight (and thus a doubly stochastic matrix), and (ii) keeping the original losses while employing a $λ$-induced row-stochastic matrix. Although prior work shows that both strategies yield the same expected descent direction for the global loss, it remains unclear whether the Euclidean-space guarantees are tight and what fundamentally differentiates their behaviors. To clarify this, we develop a weighted Hilbert-space framework $L^2(λ;\mathbb{R}^d)$ and obtain convergence rates that are strictly tighter than those from Euclidean analysis. In this geometry, the row-stochastic matrix becomes self-adjoint whereas the doubly stochastic one does not, creating additional penalty terms that amplify consensus error, thereby slowing convergence. Consequently, the difference in convergence arises not only from spectral gaps but also from these penalty terms. We then derive sufficient conditions under which the row-stochastic design converges faster even with a smaller spectral gap. Finally, by using a Rayleigh-quotient and Loewner-order eigenvalue comparison, we further obtain topology conditions that guarantee this advantage and yield practical topology-design guidelines.

Shaojie Bai, Mohammad Sadegh Talebi, Chengcheng Zhao et al.

Reinforcement learning (RL) is a powerful tool for sequential decision-making, but its application is often hindered by privacy concerns arising from its interaction data. This challenge is particularly acute in advanced networked systems, where learning from operational and user data can expose systems to privacy inference attacks. Existing differential privacy (DP) models for RL are often inadequate: the centralized model requires a fully trusted server, creating a single point of failure risk, while the local model incurs significant performance degradation that is unsuitable for many networked applications. This paper addresses this gap by leveraging the emerging shuffle model of privacy, an intermediate trust model that provides strong privacy guarantees without a centralized trust assumption. We present Shuffle Differentially Private Policy Elimination (SDP-PE), the first generic policy elimination-based algorithm for episodic RL under the shuffle model. Our method introduces a novel exponential batching schedule and a ``forgetting'' mechanism to balance the competing demands of privacy and learning performance. Our analysis shows that SDP-PE achieves a near-optimal regret bound, demonstrating a superior privacy-regret trade-off with utility comparable to the centralized model while significantly outperforming the local model. The numerical experiments also corroborate our theoretical results and demonstrate the effectiveness of SDP-PE. This work establishes the viability of the shuffle model for secure data-driven decision-making in networked systems.