Reinforcement learning (RL) is proven effective in optimizing home battery charging coordination within smart grids. However, its vulnerability to adversarial behavior poses a significant challenge to the security and fairness of the charging process. In this study, we, first, craft five stealthy false data injection (FDI) attacks that under-report the state-of-charge (SoC) values to deceive the RL agent into prioritizing their charging requests, and then, we investigate the impact of these attacks on the charging coordination system. Our evaluations demonstrate that attackers can increase their chances of charging compared to honest consumers. As a result, honest consumers experience reduced charging levels for their batteries, leading to a degradation in the system’s performance in terms of fairness, consumer satisfaction, and overall reward. These negative effects become more severe as the amount of power allocated for charging decreases and as the number of attackers in the system increases. Since the total available power for charging is limited, some honest consumers with genuinely low SoC values are not selected, creating a significant disparity in battery charging levels between honest and malicious consumers. To counter this serious threat, we develop a deep learning-based FDI attack detector and evaluated it using a real-world dataset. Our experiments show that our detector can identify malicious consumers with high accuracy and low false alarm rates, effectively protecting the RL-based charging coordination system from FDI attacks and mitigating the negative impacts of these attacks.

Recent research has shown promise in using machine learning for cyberattack detection in power systems. However, current studies face limitations: (a) dependence on either physical or cyber features, overlooking multi-modal cyber-physical (CP) correlations; (b) unrealistic full observability assumptions; (c) focus on detecting basic attacks instead of advanced threats such as ransomware (RW); and (d) use of deep learning (DL) models built for 2D data, despite the graph-structured nature of power systems. To address these gaps, we develop a CP testbed using OPAL-RT and a cyber range to simulate both physical and cyber layers under full and partial observability. The testbed produces a realistic multi-modal dataset covering normal operations and various cyberattacks, including RW, brute force, false data injection, reverse shell, and backdoor. Using this dataset, we design graph neural network (GNN)-based multi-modal intrusion detection systems (IDSs) that fuse CP features and capture spatio-temporal dependencies. Results show that CP fusion improves detection rates (DRs) by up to 16% compared to single-modal inputs. The proposed GNN-based IDSs outperform benchmarks by up to 26% in DR, remain effective under partial observability, and demonstrate up to 6% improvement in scalability when applied to larger system topologies.

The integration of renewable energy sources, such as rooftop solar panels, into smart grids poses significant challenges for managing customer-side battery storage. In response, this paper introduces a novel reinforcement learning (RL) approach aimed at optimizing the coordination of these batteries. Our approach utilizes a single-agent, multi-environment RL system designed to balance power saving, customer satisfaction, and fairness in power distribution. The RL agent dynamically allocates charging power while accounting for individual battery levels and grid constraints, employing an actor–critic algorithm. The actor determines the optimal charging power based on real-time conditions, while the critic iteratively refines the policy to enhance overall performance. The key advantages of our approach include: (1) Adaptive Power Allocation: The RL agent effectively reduces overall power consumption by optimizing grid power allocation, leading to more efficient energy use. (2) Enhanced Customer Satisfaction: By increasing the total available power from the grid, our approach significantly reduces instances of battery levels falling below the critical state of charge (SoC), thereby improving customer satisfaction. (3) Fair Power Distribution: Fairness improvements are notable, with the highest fair reward rising by 173.7% across different scenarios, demonstrating the effectiveness of our method in minimizing discrepancies in power distribution. (4) Improved Total Reward: The total reward also shows a significant increase, up by 94.1%, highlighting the efficiency of our RL-based approach. Experimental results using a real-world dataset confirm that our RL approach markedly improves fairness, power efficiency, and customer satisfaction, underscoring its potential for optimizing smart grid operations and energy management systems.

Smart power grids (SGs) enhance efficiency, reliability, and sustainability by integrating distributed energy resources (DERs) such as solar panels and wind turbines. A key challenge in SGs is managing home battery charging during periods of insufficient renewable energy generation to ensure fairness, efficiency, and customer satisfaction. This paper introduces a secure reinforcement learning (RL)-based framework for optimizing battery charging coordination while addressing privacy concerns and false data injection (FDI) attacks. Privacy is preserved through Federated Learning (FL), enabling collaborative model training without sharing sensitive State of Charge (SoC) data that could reveal personal routines. To combat FDI attacks, Deep Learning (DL)-based detectors are deployed to identify malicious SoC data manipulation. To improve FL efficiency, the Change and Transmit (CAT) technique reduces communication overhead by transmitting only model parameters that experience enough change comparing to the last round. Extensive experiments validate the framework’s efficacy. The RL-based charging coordinator ensures fairness by maintaining SoC levels within thresholds and reduces overall power utilization through optimal grid power allocation. The CAT-FL approach achieves up to 93.5% communication overhead reduction, while DL-based detectors maintain high accuracy, with supervised models reaching 99.84% and anomaly detection models achieving 92.1%. Moreover, the RL agent trained via FL demonstrates strong generalization, achieving high cumulative rewards and equitable power allocation when applied to new data owners which did not participate in FL training. This framework provides a scalable, privacy-preserving, and efficient solution for energy management in SGs, offering high accuracy against FDI attacks and paving the way for the future of smart grid deployments.

Uncrewed aerial vehicle (UAV) swarms’ ability to dynamically adapt to communication topologies brings both great potential and significant security risks. Traditional intrusion detection systems (IDSs) are designed for UAVs with fixed swarm topology and fail to secure dynamic topologies with constantly evolving communication patterns. The lack of datasets that reflect swarm behaviors across varying topologies further hinders the development of IDS. Moreover, these IDS typically focus on temporal information, overlooking the crucial spatial relationships within the UAV swarm. To address these challenges, we developed a testbed of six UAVs configured to communicate in six distinct topological graphs where each UAV acts as a node and communicates with its immediate neighbors. We then executed various cyberattacks, such as false data injection (FDI), evil twin, replay, and denial-of-service (DoS) attacks, and collected the data under both normal and attack conditions. We propose a graph neural network (GNN)-based IDS that explicitly incorporates spatial and temporal information patterns to detect intrusions. This article seeks to answer the following questions: 1) Does exploiting spatiotemporal correlations improve detection compared to IDS trained solely on temporal data? 2) Can we develop a generalized IDS capable of detecting intrusions irrespective of the specific swarm communication topology? 3) Does the spatiotemporal correlation improve detection capabilities when the range of swarm topological data is expanded in training and the IDS is tested on unseen/new topology? Through extensive experiments on varying swarm topologies and comparison with traditional deep neural network models, we assess the effectiveness of the topology-aware GNN-based IDS in securing UAV swarm communications.