# Dropout-Robust Mechanisms for Differentially Private and Fully Decentralized Mean Estimation

arXiv: https://arxiv.org/abs/2506.03746 · HTML: https://mlsanity.com/paper/2506.03746/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 29 of its 72 references that are in the map.
- First posted on arXiv 4 Jun 2025.

## Builds on (most cited first, 29 of 29)

- [Communication-Efficient Learning of Deep Networks from Decentralized Data](https://mlsanity.com/paper/1602.05629/) (2016)
- [Inverting Gradients -- How easy is it to break privacy in federated learning?](https://mlsanity.com/paper/2003.14053/) (2020)
- [Renyi Differential Privacy](https://mlsanity.com/paper/1702.07476/) (2017)
- [Exploiting Unintended Feature Leakage in Collaborative Learning](https://mlsanity.com/paper/1805.04049/) (2018)
- [RAPPOR: Randomized Aggregatable Privacy-Preserving Ordinal Response](https://mlsanity.com/paper/1407.6981/) (2014)
- [The Distributed Discrete Gaussian Mechanism for Federated Learning with Secure Aggregation](https://mlsanity.com/paper/2102.06387/) (2021)
- [Distributed Differential Privacy via Shuffling](https://mlsanity.com/paper/1808.01394/) (2018)
- [Turbo-Aggregate: Breaking the Quadratic Aggregation Barrier in Secure Federated Learning](https://mlsanity.com/paper/2002.04156/) (2020)
- [Slow and Stale Gradients Can Win the Race: Error-Runtime Trade-offs in Distributed SGD](https://mlsanity.com/paper/1803.01113/) (2018)
- [Breaking the Communication-Privacy-Accuracy Trilemma](https://mlsanity.com/paper/2007.11707/) (2020)
- [The Skellam Mechanism for Differentially Private Federated Learning](https://mlsanity.com/paper/2110.04995/) (2021)
- [MATCHA: Speeding Up Decentralized SGD via Matching Decomposition Sampling](https://mlsanity.com/paper/1905.09435/) (2019)
- [Discrete Distribution Estimation under Local Privacy](https://mlsanity.com/paper/1602.07387/) (2016)
- [Private Summation in the Multi-Message Shuffle Model](https://mlsanity.com/paper/2002.00817/) (2020)
- [ABIDES: Towards High-Fidelity Market Simulation for AI Research](https://mlsanity.com/paper/1904.12066/) (2019)
- [Muffliato: Peer-to-Peer Privacy Amplification for Decentralized Optimization and Averaging](https://mlsanity.com/paper/2206.05091/) (2022)
- [Local Privacy, Data Processing Inequalities, and Statistical Minimax Rates](https://mlsanity.com/paper/1302.3203/) (2013)
- [On the (In)security of Peer-to-Peer Decentralized Machine Learning](https://mlsanity.com/paper/2205.08443/) (2022)
- [An Accurate, Scalable and Verifiable Protocol for Federated Differentially Private Averaging](https://mlsanity.com/paper/2006.07218/) (2020)
- [Model Accuracy and Runtime Tradeoff in Distributed Deep Learning:A Systematic Study](https://mlsanity.com/paper/1509.04210/) (2015)
- [The Privacy Power of Correlated Noise in Decentralized Learning](https://mlsanity.com/paper/2405.01031/) (2024)
- [Private Counting from Anonymous Messages: Near-Optimal Accuracy with Vanishing Communication Overhead](https://mlsanity.com/paper/2106.04247/) (2021)
- [Accelerated Gossip in Networks of Given Dimension using Jacobi Polynomial Iterations](https://mlsanity.com/paper/1805.08531/) (2018)
- [Privacy Attacks in Decentralized Learning](https://mlsanity.com/paper/2402.10001/) (2024)
- [Differentially private and decentralized randomized power method](https://mlsanity.com/paper/2411.01931/) (2024)
- [Correlated Privacy Mechanisms for Differentially Private Distributed Mean Estimation](https://mlsanity.com/paper/2407.03289/) (2024)
- [Distributed Differentially Private Computation of Functions with Correlated Noise](https://mlsanity.com/paper/1904.10059/) (2019)
- [Differentially Private Empirical Cumulative Distribution Functions](https://mlsanity.com/paper/2502.06651/) (2025)
- [On Using Secure Aggregation in Differentially Private Federated Learning with Multiple Local Steps](https://mlsanity.com/paper/2407.19286/) (2024)

## Abstract (arXiv)

Achieving differentially private computations in decentralized settings poses significant challenges, particularly regarding accuracy, communication cost, and robustness against information leakage. While cryptographic solutions offer promise, they often suffer from high communication overhead or require centralization in the presence of network failures. Conversely, existing fully decentralized approaches typically rely on relaxed adversarial models or pairwise noise cancellation, the latter suffering from substantial accuracy degradation if parties unexpectedly disconnect. In this work, we propose IncA, a new protocol for fully decentralized mean estimation, a widely used primitive in data-intensive processing. Our protocol, which enforces differential privacy, requires no central orchestration and employs low-variance correlated noise, achieved by incrementally injecting sensitive information into the computation. First, we theoretically demonstrate that, when no parties permanently disconnect, our protocol achieves accuracy comparable to that of a centralized setting-already an improvement over most existing decentralized differentially private techniques. Second, we empirically show that our use of low-variance correlated noise significantly mitigates the accuracy loss experienced by existing techniques in the presence of dropouts.