kth.sePublications
Change search
Link to record
Permanent link

Direct link
Alternative names
Publications (10 of 966) Show all publications
Liu, C., Johansson, K. H. & Shi, Y. (2024). Distributed empirical risk minimization with differential privacy. Automatica, 162, Article ID 111514.
Open this publication in new window or tab >>Distributed empirical risk minimization with differential privacy
2024 (English)In: Automatica, ISSN 0005-1098, E-ISSN 1873-2836, Vol. 162, article id 111514Article in journal (Refereed) Published
Abstract [en]

This work studies the distributed empirical risk minimization (ERM) problem under differential privacy (DP) constraint. Standard distributed algorithms achieve DP typically by perturbing all local subgradients with noise, leading to significantly degenerated utility. To tackle this issue, we develop a class of private distributed dual averaging (DDA) algorithms, which activates a fraction of nodes to perform optimization. Such subsampling procedure provably amplifies the DP guarantee, thereby achieving an equivalent level of DP with reduced noise. We prove that the proposed algorithms have utility loss comparable to centralized private algorithms for both general and strongly convex problems. When removing the noise, our algorithm attains the optimal O(1/t) convergence for non-smooth stochastic optimization. Finally, experimental results on two benchmark datasets are given to verify the effectiveness of the proposed algorithms.

Place, publisher, year, edition, pages
Elsevier BV, 2024
Keywords
Differential privacy, Distributed optimization, Dual averaging, Empirical risk minimization
National Category
Control Engineering
Identifiers
urn:nbn:se:kth:diva-342825 (URN)10.1016/j.automatica.2024.111514 (DOI)001170781200001 ()2-s2.0-85182977636 (Scopus ID)
Note

QC 20240201

Available from: 2024-01-31 Created: 2024-01-31 Last updated: 2024-03-18Bibliographically approved
Ruggeri, F., Terra, A., Inam, R. & Johansson, K. H. (2024). Evaluation of Intrinsic Explainable Reinforcement Learning in Remote Electrical Tilt Optimization. In: Proceedings of 8th International Congress on Information and Communication Technology - ICICT 2023: . Paper presented at 8th International Congress on Information and Communication Technology, ICICT 2023, London, United Kingdom of Great Britain and Northern Ireland, Feb 20 2023 - Feb 23 2023 (pp. 835-854). Springer Nature
Open this publication in new window or tab >>Evaluation of Intrinsic Explainable Reinforcement Learning in Remote Electrical Tilt Optimization
2024 (English)In: Proceedings of 8th International Congress on Information and Communication Technology - ICICT 2023, Springer Nature , 2024, p. 835-854Conference paper, Published paper (Refereed)
Abstract [en]

This paper empirically evaluates two intrinsic Explainable Reinforcement Learning (XRL) algorithms on the Remote Electrical Tilt (RET) optimization problem. In RET optimization, where the electrical downtilt of the antennas in a cellular network is controlled to optimize coverage and capacity, explanations are necessary to understand the reasons behind a specific adjustment. First, we formulate the RET problem in the reinforcement learning (RL) framework and describe how we apply Decomposed Reward Deep Q Network (drDQN) and Linear ModelU-Tree (LMUT), which are two state-of-the-art XRL algorithms. Then, we train and test such agents in a realistic simulated network. Our results highlight both advantages and disadvantages of the algorithms. DrDQN provides intuitive contrastive local explanations for the agent’s decisions to adjust the downtilt of an antenna, while achieving the same performance as the original DQN algorithm. LMUT reaches high performance while employing a fully transparent linear model capable of generating both local and global explanations. On the other hand, drDQN adds a constraint on the reward design that might be problematic for the specification of the objective, whereas LMUT could generate misleading global feature importance and needs additional developments to provide more user-interpretable local explanations.

Place, publisher, year, edition, pages
Springer Nature, 2024
Keywords
Artificial intelligence, Cellular networks, Explainable reinforcement learning, Reinforcement learning, Remote electrical tilt optimization
National Category
Computer Sciences
Identifiers
urn:nbn:se:kth:diva-339274 (URN)10.1007/978-981-99-3236-8_67 (DOI)2-s2.0-85174736754 (Scopus ID)
Conference
8th International Congress on Information and Communication Technology, ICICT 2023, London, United Kingdom of Great Britain and Northern Ireland, Feb 20 2023 - Feb 23 2023
Note

Part of ISBN 9789819932351

QC 20231106

Available from: 2023-11-06 Created: 2023-11-06 Last updated: 2023-11-06Bibliographically approved
Rikos, A. I., Hadjicostis, C. N. & Johansson, K. H. (2024). Finite time quantized average consensus with transmission stopping guarantees and no quantization error. Automatica, 163, Article ID 111522.
Open this publication in new window or tab >>Finite time quantized average consensus with transmission stopping guarantees and no quantization error
2024 (English)In: Automatica, ISSN 0005-1098, E-ISSN 1873-2836, Vol. 163, article id 111522Article in journal (Refereed) Published
Abstract [en]

Networked control systems, which are composed of spatially distributed sensors and actuators that communicate through wireless networks, are emerging as a fundamental infrastructure technology in 5G and IoT technologies. In order to increase flexibility and reduce deployment and maintenance costs, their operation needs to guarantee (i) efficient communication between nodes and (ii) preservation of available energy. Motivated by these requirements, we present and analyze a novel distributed average consensus algorithm, which (i) operates exclusively on quantized values (in order to guarantee efficient communication and data storage), (ii) relies on event-driven updates (in order to reduce energy consumption, communication bandwidth, network congestion, and/or processor usage), and (iii) allows each node to cease transmissions once the exact average of the initial quantized values has been reached (in order to preserve its stored energy). We characterize the properties of the proposed algorithm and show that its execution, on any time-invariant and strongly connected digraph, allows all nodes to reach in finite time a common consensus value that is equal to the exact average (represented as the ratio of two quantized values). Then, we present upper bounds on (i) the number of transmissions and computations each node has to perform during the execution of the algorithm, and (ii) the memory and energy requirements of each node in order for the algorithm to be executed. Finally, we provide examples that demonstrate the operation, performance, and potential advantages of our proposed algorithm.

Place, publisher, year, edition, pages
Elsevier BV, 2024
Keywords
Digraphs, Event-triggered distributed algorithms, Multi-agent systems, Quantization, Quantized average consensus
National Category
Control Engineering
Identifiers
urn:nbn:se:kth:diva-343663 (URN)10.1016/j.automatica.2024.111522 (DOI)2-s2.0-85184838548 (Scopus ID)
Note

QC 20240222

Available from: 2024-02-22 Created: 2024-02-22 Last updated: 2024-02-22Bibliographically approved
Yang, Z., Fonseca, J., Zhu, S., Chen, C., Guan, X. & Johansson, K. H. (2023). Adaptive Estimation for Environmental Monitoring Using an Autonomous Underwater Vehicle. In: 2023 62nd IEEE Conference on Decision and Control, CDC 2023: . Paper presented at 62nd IEEE Conference on Decision and Control, CDC 2023, Singapore, Singapore, Dec 13 2023 - Dec 15 2023 (pp. 2521-2528). Institute of Electrical and Electronics Engineers (IEEE)
Open this publication in new window or tab >>Adaptive Estimation for Environmental Monitoring Using an Autonomous Underwater Vehicle
Show others...
2023 (English)In: 2023 62nd IEEE Conference on Decision and Control, CDC 2023, Institute of Electrical and Electronics Engineers (IEEE) , 2023, p. 2521-2528Conference paper, Published paper (Refereed)
Abstract [en]

This paper considers the problem of monitoring and adaptively estimating an environmental field, such as temperature or salinity, using an autonomous underwater vehicle (AUV). The AUV moves in the field and persistently measures environmental scalars and its position in its local coordinate frame. The environmental scalars are approximately linearly distributed over the region of interest, and an adaptive estimator is designed to estimate the gradient. By orthogonal decomposition of the velocity of the AUV, a linear time-varying system is equivalently constructed, and the sufficient conditions on the motion of the AUV are established, under which the global exponential stability of the estimation error system is rigorously proved. Furthermore, an estimate of the exponential convergence rate is given, and a reference trajectory that maximizes the estimate of the convergence rate is obtained for the AUV to track. Numerical examples verify the stability and efficiency of the system.

Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers (IEEE), 2023
Keywords
Adaptive Estimation, Autonomous Underwater Vehicle, Environmental Monitoring, Exponential Stability
National Category
Control Engineering
Identifiers
urn:nbn:se:kth:diva-343722 (URN)10.1109/CDC49753.2023.10383773 (DOI)2-s2.0-85184828866 (Scopus ID)
Conference
62nd IEEE Conference on Decision and Control, CDC 2023, Singapore, Singapore, Dec 13 2023 - Dec 15 2023
Note

Part of proceedings ISBN 9798350301243

QC 20240222

Available from: 2024-02-22 Created: 2024-02-22 Last updated: 2024-02-29Bibliographically approved
Fonseca, J., Rocha, A., Aguiar, M. & Johansson, K. H. (2023). Adaptive Sampling of Algal Blooms Using an Autonomous Underwater Vehicle and Satellite Imagery. In: 2023 IEEE Conference on Control Technology and Applications, CCTA 2023: . Paper presented at 2023 IEEE Conference on Control Technology and Applications, CCTA 2023, Bridgetown, Barbados, Aug 16 2023 - Aug 18 2023 (pp. 638-644). Institute of Electrical and Electronics Engineers (IEEE)
Open this publication in new window or tab >>Adaptive Sampling of Algal Blooms Using an Autonomous Underwater Vehicle and Satellite Imagery
2023 (English)In: 2023 IEEE Conference on Control Technology and Applications, CCTA 2023, Institute of Electrical and Electronics Engineers (IEEE) , 2023, p. 638-644Conference paper, Published paper (Refereed)
Abstract [en]

This paper proposes a method that uses satellite data to improve adaptive sampling missions. We find and track algal bloom fronts using an autonomous underwater vehicle (AUV) equipped with a sensor that measures the concentration of chlorophyll a. Chlorophyll a concentration indicates the presence of algal blooms. The proposed method learns the kernel parameters of a Gaussian process model using satellite images of chlorophyll a from previous days. The AUV estimates the chlorophyll a concentration online using locally collected data. The algal bloom front estimate is fed to the motion control algorithm. The performance of this method is evaluated through simulations using a real dataset of an algal bloom front in the Baltic. We consider a real-world scenario with sensor and localization noise and with a detailed AUV model.

Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers (IEEE), 2023
National Category
Oceanography, Hydrology and Water Resources
Identifiers
urn:nbn:se:kth:diva-338992 (URN)10.1109/CCTA54093.2023.10252251 (DOI)2-s2.0-85173889475 (Scopus ID)
Conference
2023 IEEE Conference on Control Technology and Applications, CCTA 2023, Bridgetown, Barbados, Aug 16 2023 - Aug 18 2023
Note

Part of ISBN 9798350335446

QC 20231123

Available from: 2023-11-01 Created: 2023-11-01 Last updated: 2023-11-23Bibliographically approved
Xing, Y. & Johansson, K. H. (2023). Almost Exact Recovery in Gossip Opinion Dynamics Over Stochastic Block Models. In: 2023 62nd IEEE Conference on Decision and Control, CDC 2023: . Paper presented at 62nd IEEE Conference on Decision and Control, CDC 2023, Singapore, Singapore, Dec 13 2023 - Dec 15 2023 (pp. 2421-2426). Institute of Electrical and Electronics Engineers (IEEE)
Open this publication in new window or tab >>Almost Exact Recovery in Gossip Opinion Dynamics Over Stochastic Block Models
2023 (English)In: 2023 62nd IEEE Conference on Decision and Control, CDC 2023, Institute of Electrical and Electronics Engineers (IEEE) , 2023, p. 2421-2426Conference paper, Published paper (Refereed)
Abstract [en]

We study community detection based on state observations from gossip opinion dynamics over stochastic block models (SBM). It is assumed that a network is generated from a two-community SBM where each agent has a community label and each edge exists with probability depending on its endpoints' labels. A gossip process then evolves over the sampled network. We propose two algorithms to detect the communities out of a single trajectory of the process. It is shown that, when the influence of stubborn agents is small and the link probability within communities is large, an algorithm based on clustering transient agent states can achieve almost exact recovery of the communities. That is, the algorithm can recover all but a vanishing part of community labels with high probability. In contrast, when the influence of stubborn agents is large, another algorithm based on clustering time average of agent states can achieve almost exact recovery. Numerical experiments are given for illustration of the two algorithms and the theoretical results of the paper.

Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers (IEEE), 2023
National Category
Control Engineering
Identifiers
urn:nbn:se:kth:diva-343716 (URN)10.1109/CDC49753.2023.10383465 (DOI)2-s2.0-85184803048 (Scopus ID)
Conference
62nd IEEE Conference on Decision and Control, CDC 2023, Singapore, Singapore, Dec 13 2023 - Dec 15 2023
Note

Part of ISBN 9798350301243

QC 20240222

Available from: 2024-02-22 Created: 2024-02-22 Last updated: 2024-02-22Bibliographically approved
He, X., Yi, X., Zhao, Y., Johansson, K. H. & Gupta, V. (2023). Asymptotic Analysis of Federated Learning Under Event-Triggered Communication. IEEE Transactions on Signal Processing, 71, 2654-2667
Open this publication in new window or tab >>Asymptotic Analysis of Federated Learning Under Event-Triggered Communication
Show others...
2023 (English)In: IEEE Transactions on Signal Processing, ISSN 1053-587X, E-ISSN 1941-0476, Vol. 71, p. 2654-2667Article in journal (Refereed) Published
Abstract [en]

Federated learning (FL) is a collaborative machine learning (ML) paradigm based on persistent communication between a central server and multiple edge devices. However, frequent communication of large ML models can incur considerable communication resources, especially costly for wireless network nodes. In this paper, we develop a communication-efficient protocol to reduce the number of communication instances in each round while maintaining convergence rate and asymptotic distribution properties. First, we propose a novel communication-efficient FL algorithm that utilizes an event-triggered communication mechanism, where each edge device updates the model by using stochastic gradient descent with local sampling data and the central server aggregates all local models from the devices by using model averaging. Communication can be reduced since each edge device and the central server transmits its updated model only when the difference between the current model and the last communicated model is larger than a threshold. Thresholds of the devices and server are not necessarily coordinated, and the thresholds and step sizes are not constrained to be of specific forms. Under mild conditions on loss functions, step sizes and thresholds, for the proposed algorithm, we establish asymptotic analysis results in three ways, respectively: convergence in expectation, almost-sure convergence, and asymptotic distribution of the estimation error. In addition, we show that by fine-tunning the parameters, the proposed event-triggered FL algorithm can reach the same convergence rate as state-of-the-art results from existing time-driven FL. We also establish asymptotic efficiency in the sense of Central Limit Theorem of the estimation error. Numerical simulations for linear regression and image classification problems in the literature are provided to show the effectiveness of the developed results.

Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers (IEEE), 2023
Keywords
Federated learning, asymptotic convergence, event-triggered, stochastic gradient descent, distributed optimization
National Category
Control Engineering
Identifiers
urn:nbn:se:kth:diva-335203 (URN)10.1109/TSP.2023.3295734 (DOI)001049952000001 ()2-s2.0-85165250225 (Scopus ID)
Note

QC 20230914

Available from: 2023-09-14 Created: 2023-09-14 Last updated: 2023-09-14Bibliographically approved
Rikos, A. I., Jiang, W., Charalambous, T. & Johansson, K. H. (2023). Asynchronous Distributed Optimization via ADMM with Efficient Communication. In: 2023 62nd IEEE Conference on Decision and Control, CDC 2023: . Paper presented at 62nd IEEE Conference on Decision and Control, CDC 2023, Singapore, Singapore, Dec 13 2023 - Dec 15 2023 (pp. 7002-7008). Institute of Electrical and Electronics Engineers (IEEE)
Open this publication in new window or tab >>Asynchronous Distributed Optimization via ADMM with Efficient Communication
2023 (English)In: 2023 62nd IEEE Conference on Decision and Control, CDC 2023, Institute of Electrical and Electronics Engineers (IEEE) , 2023, p. 7002-7008Conference paper, Published paper (Refereed)
Abstract [en]

In this paper, we focus on an asynchronous distributed optimization problem. In our problem, each node is endowed with a convex local cost function, and is able to communicate with its neighbors over a directed communication network. Furthermore, we assume that the communication channels between nodes have limited bandwidth, and each node suffers from processing delays. We present a distributed algorithm which combines the Alternating Direction Method of Multipliers (ADMM) strategy with a finite time quantized averaging algorithm. In our proposed algorithm, nodes exchange quantized valued messages and operate in an asynchronous fashion. More specifically, during every iteration of our algorithm each node (i) solves a local convex optimization problem (for the one of its primal variables), and (ii) utilizes a finite-time quantized averaging algorithm to obtain the value of the second primal variable (since the cost function for the second primal variable is not decomposable). We show that our algorithm converges to the optimal solution at a rate of O (1/ k) (where k is the number of time steps) for the case where the local cost function of every node is convex and not-necessarily differentiable. Finally, we demonstrate the operational advantages of our algorithm against other algorithms from the literature.

Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers (IEEE), 2023
National Category
Control Engineering
Identifiers
urn:nbn:se:kth:diva-343737 (URN)10.1109/CDC49753.2023.10383681 (DOI)2-s2.0-85184819753 (Scopus ID)
Conference
62nd IEEE Conference on Decision and Control, CDC 2023, Singapore, Singapore, Dec 13 2023 - Dec 15 2023
Note

Part of ISBN 9798350301243

QC 20240222

Available from: 2024-02-22 Created: 2024-02-22 Last updated: 2024-02-29Bibliographically approved
Mavridis, C. N., Kanellopoulos, A., Vamvoudakis, K. G., Baras, J. S. & Johansson, K. H. (2023). Attack Identification for Cyber-Physical Security in Dynamic Games under Cognitive Hierarchy. In: Ifac-Papersonline: . Paper presented at 22nd IFAC World Congress, Yokohama, Japan, Jul 9-14 2023 (pp. 11223-11228). Elsevier BV
Open this publication in new window or tab >>Attack Identification for Cyber-Physical Security in Dynamic Games under Cognitive Hierarchy
Show others...
2023 (English)In: Ifac-Papersonline, Elsevier BV , 2023, p. 11223-11228Conference paper, Published paper (Refereed)
Abstract [en]

This paper considers the problem of identifying the profiles and capabilities of attackers injecting adversarial inputs to a cyber-physical system. The system in question interacts with attackers of different levels of intelligence, each employing different feedback controllers against the system. Principles of behavioral game theory - specifically the concept of level-k thinking - is employed to construct a database of potential attack vectors. By observing the state trajectories under sequential interactions with different adversaries, the defender adaptively estimates both the number and profiles of the different attack signals using an online deterministic annealing approach. This information is used to dynamically estimate the level of intelligence of the attackers. Simulation results showcase the efficacy of the proposed method.

Place, publisher, year, edition, pages
Elsevier BV, 2023
National Category
Control Engineering
Identifiers
urn:nbn:se:kth:diva-343703 (URN)10.1016/j.ifacol.2023.10.851 (DOI)2-s2.0-85184825580 (Scopus ID)
Conference
22nd IFAC World Congress, Yokohama, Japan, Jul 9-14 2023
Note

Part of ISBN: 978-171387234-4

QC 20240223

Available from: 2024-02-22 Created: 2024-02-22 Last updated: 2024-02-23Bibliographically approved
Yi, X., Zhang, S., Yang, T., Chai, T. & Johansson, K. H. (2023). Communication Compression for Distributed Nonconvex Optimization. IEEE Transactions on Automatic Control, 68(9), 5477-5492
Open this publication in new window or tab >>Communication Compression for Distributed Nonconvex Optimization
Show others...
2023 (English)In: IEEE Transactions on Automatic Control, ISSN 0018-9286, E-ISSN 1558-2523, Vol. 68, no 9, p. 5477-5492Article in journal (Refereed) Published
Abstract [en]

This paper considers distributed nonconvex optimization with the cost functions being distributed over agents. Noting that information compression is a key tool to reduce the heavy communication load for distributed algorithms as agents iteratively communicate with neighbors, we propose three distributed primal–dual algorithms with compressed communication. The first two algorithms are applicable to a general class of compressors with bounded relative compression error and the third algorithm is suitable for two general classes of compressors with bounded absolute compression error. We show that the proposed distributed algorithms with compressed communication have comparable convergence properties as state-of-the-art algorithms with exact communication. Specifically, we show that they can find first-order stationary points with sublinear convergence rate when each local cost function is smooth, where is the total number of iterations, and find global optima with linear convergence rate under an additional condition that the global cost function satisfies the Polyak–Łojasiewicz condition. Numerical simulations are provided to illustrate the effectiveness of the theoretical results.

Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers (IEEE), 2023
Keywords
Communication compression, distributed optimization, linear convergence, nonconvex optimization, Polyak–Łojasiewicz condition
National Category
Control Engineering
Identifiers
urn:nbn:se:kth:diva-335765 (URN)10.1109/TAC.2022.3225515 (DOI)001059698200021 ()2-s2.0-85144046697 (Scopus ID)
Note

QC 20230911

Available from: 2023-09-11 Created: 2023-09-11 Last updated: 2023-10-25Bibliographically approved
Organisations
Identifiers
ORCID iD: ORCID iD iconorcid.org/0000-0001-9940-5929

Search in DiVA

Show all publications