arXiv subjects

Stefan Werner

Publications and source records attributed to Stefan Werner.

33 records · Page 2Linked to original sources

arXiv2023

Distributed Filtering Design with Enhanced Resilience to Coordinated Byzantine Attacks

This paper proposes a Byzantine-resilient consensus-based distributed filter (BR-CDF) wherein network agents employ partial sharing of state parameters. We characterize the performance and convergence of the BR-CDF and study the impact of a coordinated data falsification attack. Our analysis shows that sharing merely a fraction of the states improves robustness against coordinated Byzantine attacks. In addition, we model the optimal attack strategy as an optimization problem where Byzantine agents design their attack covariance or the sequence of shared fractions to maximize the network-wide mean squared error (MSE). Numerical results demonstrate the accuracy of the proposed BR-CDF and its robustness against Byzantine attacks. Furthermore, the simulation results show that the influence of the covariance design is more pronounced when agents exchange larger portions of their states with neighbors. In contrast, the performance is more sensitive to the sequence of shared fractions when smaller portions are exchanged.

eess.SP↗

arXiv2023

Personalized Graph Federated Learning with Differential Privacy

This paper presents a personalized graph federated learning (PGFL) framework in which distributedly connected servers and their respective edge devices collaboratively learn device or cluster-specific models while maintaining the privacy of every individual device. The proposed approach exploits similarities among different models to provide a more relevant experience for each device, even in situations with diverse data distributions and disproportionate datasets. Furthermore, to ensure a secure and efficient approach to collaborative personalized learning, we study a variant of the PGFL implementation that utilizes differential privacy, specifically zero-concentrated differential privacy, where a noise sequence perturbs model exchanges. Our mathematical analysis shows that the proposed privacy-preserving PGFL algorithm converges to the optimal cluster-specific solution for each cluster in linear time. It also shows that exploiting similarities among clusters leads to an alternative output whose distance to the original solution is bounded, and that this bound can be adjusted by modifying the algorithm's hyperparameters. Further, our analysis shows that the algorithm ensures local differential privacy for all clients in terms of zero-concentrated differential privacy. Finally, the performance of the proposed PGFL algorithm is examined by performing numerical experiments in the context of regression and classification using synthetic data and the MNIST dataset.

cs.LG↗

arXiv2023

Asynchronous Online Federated Learning with Reduced Communication Requirements

Online federated learning (FL) enables geographically distributed devices to learn a global shared model from locally available streaming data. Most online FL literature considers a best-case scenario regarding the participating clients and the communication channels. However, these assumptions are often not met in real-world applications. Asynchronous settings can reflect a more realistic environment, such as heterogeneous client participation due to available computational power and battery constraints, as well as delays caused by communication channels or straggler devices. Further, in most applications, energy efficiency must be taken into consideration. Using the principles of partial-sharing-based communications, we propose a communication-efficient asynchronous online federated learning (PAO-Fed) strategy. By reducing the communication overhead of the participants, the proposed method renders participation in the learning task more accessible and efficient. In addition, the proposed aggregation mechanism accounts for random participation, handles delayed updates and mitigates their effect on accuracy. We prove the first and second-order convergence of the proposed PAO-Fed method and obtain an expression for its steady-state mean square deviation. Finally, we conduct comprehensive simulations to study the performance of the proposed method on both synthetic and real-life datasets. The simulations reveal that in asynchronous settings, the proposed PAO-Fed is able to achieve the same convergence properties as that of the online federated stochastic gradient while reducing the communication overhead by 98 percent.

cs.LG↗

arXiv2023

Learning a model is paramount for sample efficiency in reinforcement learning control of PDEs

The goal of this paper is to make a strong point for the usage of dynamical models when using reinforcement learning (RL) for feedback control of dynamical systems governed by partial differential equations (PDEs). To breach the gap between the immense promises we see in RL and the applicability in complex engineering systems, the main challenges are the massive requirements in terms of the training data, as well as the lack of performance guarantees. We present a solution for the first issue using a data-driven surrogate model in the form of a convolutional LSTM with actuation. We demonstrate that learning an actuated model in parallel to training the RL agent significantly reduces the total amount of required data sampled from the real system. Furthermore, we show that iteratively updating the model is of major importance to avoid biases in the RL training. Detailed ablation studies reveal the most important ingredients of the modeling process. We use the chaotic Kuramoto-Sivashinsky equation do demonstarte our findings.

cs.LG↗

arXiv2022

Decentralized Optimization with Distributed Features and Non-Smooth Objective Functions

We develop a new consensus-based distributed algorithm for solving learning problems with feature partitioning and non-smooth convex objective functions. Such learning problems are not separable, i.e., the associated objective functions cannot be directly written as a summation of agent-specific objective functions. To overcome this challenge, we redefine the underlying optimization problem as a dual convex problem whose structure is suitable for distributed optimization using the alternating direction method of multipliers (ADMM). Next, we propose a new method to solve the minimization problem associated with the ADMM update step that does not rely on any conjugate function. Calculating the relevant conjugate functions may be hard or even unfeasible, especially when the objective function is non-smooth. To obviate computing any conjugate function, we solve the optimization problem associated with each ADMM iteration in the dual domain utilizing the block coordinate descent algorithm. Unlike the existing related algorithms, the proposed algorithm is fully distributed and does away with the conjugate of the objective function. We prove theoretically that the proposed algorithm attains the optimal centralized solution. We also confirm its network-wide convergence via simulations.

eess.SP↗

arXiv2021

Resource-Aware Asynchronous Online Federated Learning for Nonlinear Regression

Many assumptions in the federated learning literature present a best-case scenario that can not be satisfied in most real-world applications. An asynchronous setting reflects the realistic environment in which federated learning methods must be able to operate reliably. Besides varying amounts of non-IID data at participants, the asynchronous setting models heterogeneous client participation due to available computational power and battery constraints and also accounts for delayed communications between clients and the server. To reduce the communication overhead associated with asynchronous online federated learning (ASO-Fed), we use the principles of partial-sharing-based communication. In this manner, we reduce the communication load of the participants and, therefore, render participation in the learning task more accessible. We prove the convergence of the proposed ASO-Fed and provide simulations to analyze its behavior further. The simulations reveal that, in the asynchronous setting, it is possible to achieve the same convergence as the federated stochastic gradient (Online-FedSGD) while reducing the communication tenfold.

cs.LG↗

arXiv2021

Communication-Efficient Online Federated Learning Framework for Nonlinear Regression

Federated learning (FL) literature typically assumes that each client has a fixed amount of data, which is unrealistic in many practical applications. Some recent works introduced a framework for online FL (Online-Fed) wherein clients perform model learning on streaming data and communicate the model to the server; however, they do not address the associated communication overhead. As a solution, this paper presents a partial-sharing-based online federated learning framework (PSO-Fed) that enables clients to update their local models using continuous streaming data and share only portions of those updated models with the server. During a global iteration of PSO-Fed, non-participant clients have the privilege to update their local models with new data. Here, we consider a global task of kernel regression, where clients use a random Fourier features-based kernel LMS on their data for local learning. We examine the mean convergence of the PSO-Fed for kernel regression. Experimental results show that PSO-Fed can achieve competitive performance with a significantly lower communication overhead than Online-Fed.

cs.LG↗

arXiv2021

On Stability and Convergence of Distributed Filters

Recent years have bore witness to the proliferation of distributed filtering techniques, where a collection of agents communicating over an ad-hoc network aim to collaboratively estimate and track the state of a system. These techniques form the enabling technology of modern multi-agent systems and have gained great importance in the engineering community. Although most distributed filtering techniques come with a set of stability and convergence criteria, the conditions imposed are found to be unnecessarily restrictive. The paradigm of stability and convergence in distributed filtering is revised in this manuscript. Accordingly, a general distributed filter is constructed and its estimation error dynamics is formulated. The conducted analysis demonstrates that conditions for achieving stable filtering operations are the same as those required in the centralized filtering setting. Finally, the concepts are demonstrated in a Kalman filtering framework and validated using simulation examples.

eess.SP↗

arXiv2020

Run2Survive: A Decision-theoretic Approach to Algorithm Selection based on Survival Analysis

Algorithm selection (AS) deals with the automatic selection of an algorithm from a fixed set of candidate algorithms most suitable for a specific instance of an algorithmic problem class, where "suitability" often refers to an algorithm's runtime. Due to possibly extremely long runtimes of candidate algorithms, training data for algorithm selection models is usually generated under time constraints in the sense that not all algorithms are run to completion on all instances. Thus, training data usually comprises censored information, as the true runtime of algorithms timed out remains unknown. However, many standard AS approaches are not able to handle such information in a proper way. On the other side, survival analysis (SA) naturally supports censored data and offers appropriate ways to use such data for learning distributional models of algorithm runtime, as we demonstrate in this work. We leverage such models as a basis of a sophisticated decision-theoretic approach to algorithm selection, which we dub Run2Survive. Moreover, taking advantage of a framework of this kind, we advocate a risk-averse approach to algorithm selection, in which the avoidance of a timeout is given high priority. In an extensive experimental study with the standard benchmark ASlib, our approach is shown to be highly competitive and in many cases even superior to state-of-the-art AS approaches.

cs.LG↗

arXiv2020

Secure Boot from Non-Volatile Memory for Programmable SoC Architectures

In modern embedded systems, the trust in comprehensive security standards all along the product life cycle has become an increasingly important access-to-market requirement. However, these security standards rely on mandatory immunity assumptions such as the integrity and authenticity of an initial system configuration typically loaded from Non-Volatile Memory (NVM). This applies especially to FPGA-based Programmable System-on-Chip (PSoC) architectures, since object codes as well as configuration data easily exceed the capacity of a secure bootROM. In this context, an attacker could try to alter the content of the NVM device in order to manipulate the system. The PSoC therefore relies on the integrity of the NVM particularly at boot-time. In this paper, we propose a methodology for securely booting from an NVM in a potentially unsecure environment by exploiting the reconfigurable logic of the FPGA. Here, the FPGA serves as a secure anchor point by performing required integrity and authenticity verifications prior to the configuration and execution of any user application loaded from the NVM on the PSoC. The proposed secure boot process is based on the following assumptions and steps: 1) The boot configurationis stored on a fully encrypted Secure Digital memory card (SD card) or alternatively Flash acting as NVM. 2) At boot time, a hardware design called Trusted Memory-Interface Unit (TMIU) is loaded to verify first the authenticity of the deployed NVM and then after decryption the integrity of its content. To demonstrate the practicability of our approach, we integrated the methodology into the vendor-specific secure boot process of a Xilinx Zynq PSoC and evaluated the design objectives performance, power and resource costs.

cs.CR↗

arXiv2016

On the Asymptotic Bias of the Diffusion-Based Distributed Pareto Optimization

We revisit the asymptotic bias analysis of the distributed Pareto optimization algorithm developed based on the diffusion strategies. We propose an alternative way to analyze the asymptotic bias of this algorithm at small step-sizes and show that the asymptotic bias descends to zero with a linear dependence on the largest step-size parameter when this parameter is sufficiently small. In addition, through the proposed analytic approach, we provide an expression for the small-step-size asymptotic bias when a condition assumed jointly on the combination matrices and the step-sizes does not strictly hold. This is a likely scenario in practice, which has not been considered in the original paper that introduced the algorithm. Our methodology provides new insights into the inner workings of the diffusion Pareto optimization algorithm while being considerably less involved than the small-step-size asymptotic bias analysis presented in the original work. This is because we take advantage of the special eigenstructure of the composite combination matrix used in the algorithm without calling for any eigenspace decomposition or matrix inversion.

cs.DC↗

arXiv2016

Constrained Phase Noise Estimation in OFDM Using Scattered Pilots Without Decision Feedback

In this paper, we consider an OFDM radio link corrupted by oscillator phase noise in the receiver, namely the problem of estimating and compensating for the impairment. To lessen the computational burden and delay incurred onto the receiver, we estimate phase noise using only scattered pilot subcarriers, i.e., no tentative symbol decisions are used in obtaining and improving the phase noise estimate. In particular, the phase noise estimation problem is posed as an unconstrained optimization problem whose minimizer suffers from the so-called amplitude and phase estimation error. These errors arise due to receiver noise, estimation from limited scattered pilot subcarriers and estimation using a dimensionality reduction model. It is empirically shown that, at high signal-to-noise-ratios, the phase estimation error is small. To reduce the amplitude estimation error, we restrict the minimizer to be drawn from the so-called phase noise geometry set when minimizing the cost function. The resulting optimization problem is a non-convex program. However, using the S-procedure for quadratic equalities, we show that the optimal solution can be obtained by solving the convex dual problem. We also consider a less complex heuristic scheme that achieves the same objective of restricting the minimizer to the phase noise geometry set. Through simulations, we demonstrate improved coded bit-error-rate and phase noise estimation error performance when enforcing the phase noise geometry. For example, at high signal-to-noise-ratios, the probability density function of the phase noise estimation error exhibits thinner tails which results in lower bit-error-rate.

cs.IT↗

arXiv2015

On the Mean-Square Performance of the Constrained LMS Algorithm

The so-called constrained least mean-square algorithm is one of the most commonly used linear-equality-constrained adaptive filtering algorithms. Its main advantages are adaptability and relative simplicity. In order to gain analytical insights into the performance of this algorithm, we examine its mean-square performance and derive theoretical expressions for its transient and steady-state mean-square deviation. Our methodology is inspired by the principle of energy conservation in adaptive filters. Simulation results corroborate the accuracy of the derived formula.

eess.SY↗

arXiv2014

Analysis of a Reduced-Communication Diffusion LMS Algorithm

In diffusion-based algorithms for adaptive distributed estimation, each node of an adaptive network estimates a target parameter vector by creating an intermediate estimate and then combining the intermediate estimates available within its closed neighborhood. We analyze the performance of a reduced-communication diffusion least mean-square (RC-DLMS) algorithm, which allows each node to receive the intermediate estimates of only a subset of its neighbors at each iteration. This algorithm eases the usage of network communication resources and delivers a trade-off between estimation performance and communication cost. We show analytically that the RC-DLMS algorithm is stable and convergent in both mean and mean-square senses. We also calculate its theoretical steady-state mean-square deviation. Simulation results demonstrate a good match between theory and experiment.

cs.DC↗

arXiv2014

Recursive Total Least-Squares Algorithm Based on Inverse Power Method and Dichotomous Coordinate-Descent Iterations

We develop a recursive total least-squares (RTLS) algorithm for errors-in-variables system identification utilizing the inverse power method and the dichotomous coordinate-descent (DCD) iterations. The proposed algorithm, called DCD-RTLS, outperforms the previously-proposed RTLS algorithms, which are based on the line-search method, with reduced computational complexity. We perform a comprehensive analysis of the DCD-RTLS algorithm and show that it is asymptotically unbiased as well as being stable in the mean. We also find a lower bound for the forgetting factor that ensures mean-square stability of the algorithm and calculate the theoretical steady-state mean-square deviation (MSD). We verify the effectiveness of the proposed algorithm and the accuracy of the predicted steady-state MSD via simulations.

eess.SY↗