arXiv subjects

Soumyajit Dey

Publications and source records attributed to Soumyajit Dey.

At least 19 recordsLinked to original sources

arXiv2026

Safe Controller Synthesis Using Lyapunov-based Barriers for Linear Hybrid Systems with Simplex Architecture

Modern cyber-physical systems often have a two-layered design, where the primary controller is AI-enabled or an analytical controller optimising some specific cost function. If the resulting control action is perceived as unsafe, a secondary safety-focused backup controller is activated. The existing backup controller design schemes do not consider a real-time deadline for the course correction of a potentially unsafe system trajectory or constrain maximisation of the safe operating region as a synthesis criterion. This essentially implies an eventual safety guarantee over a small operating region. This paper proposes a novel design method for backup safe controllers (BSCs) that ensure invariance across the largest possible region in the safe state space, along with a guarantee for timely recovery when the system states deviate from their usual behaviour. This is the first work to synthesise safe controllers that ensure maximal safety and timely recovery while aiming at minimal resource usage by switching between BSCs with different execution rates. An online safe controller activation policy is also proposed to switch between BSCs (and the primary optimal controller) to optimise processing bandwidth for control computation. To establish the efficacy of the proposed method, we evaluate the safety and recovery time of the proposed safe controllers, as well as the activation policy, in closed loops with linear hybrid dynamical systems under budgeted bandwidth.

eess.SY↗

arXiv2026

AD$^2$: Analysis and Detection of Adversarial Threats in Visual Perception for End-to-End Autonomous Driving Systems

End-to-end autonomous driving systems have achieved significant progress, yet their adversarial robustness remains largely underexplored. In this work, we conduct a closed-loop evaluation of state-of-the-art autonomous driving agents under black-box adversarial threat models in CARLA. Specifically, we consider three representative attack vectors on the visual perception pipeline: (i) a physics-based blur attack induced by acoustic waves, (ii) an electromagnetic interference attack that distorts captured images, and (iii) a digital attack that adds ghost objects as carefully crafted bounded perturbations on images. Our experiments on two advanced agents, Transfuser and Interfuser, reveal severe vulnerabilities to such attacks, with driving scores dropping by up to 99% in the worst case, raising valid safety concerns. To help mitigate such threats, we further propose a lightweight Attack Detection model for Autonomous Driving systems (AD$^2$) based on attention mechanisms that capture spatial-temporal consistency. Comprehensive experiments across multi-camera inputs on CARLA show that our detector achieves superior detection capability and computational efficiency compared to existing approaches.

cs.CV↗

arXiv2026

Mitigating Timing-Based Attacks in Real-Time Cyber-Physical Systems

Real-time cyber-physical systems depend on deterministic task execution to guarantee safety and correctness. Unfortunately, this determinism can unintentionally expose timing information that enables adversaries to infer task execution patterns and carry out timing-based attacks targeting safety-critical control tasks. While prior defenses aim to obscure schedules through randomization or isolation, they typically neglect the implications of such modifications on closed-loop control behavior and real-time feasibility. This work studies the problem of securing real-time control workloads against timing inference attacks while explicitly accounting for both schedulability constraints and control performance requirements. We present a scheduling-based mitigation approach that introduces bounded timing perturbations to control task executions in a structured manner, reducing adversarial opportunities without violating real-time guarantees. The framework jointly considers worst-case execution behavior and the impact of execution delays on control performance, enabling the system to operate within predefined safety and performance limits. Through experimental evaluation on representative task sets and control scenarios, the proposed approach demonstrates that exposure to timing-based attacks can be significantly reduced while preserving predictable execution and acceptable control quality.

eess.SY↗

arXiv2025

Incentivizing Safer Actions in Policy Optimization for Constrained Reinforcement Learning

Constrained Reinforcement Learning (RL) aims to maximize the return while adhering to predefined constraint limits, which represent domain-specific safety requirements. In continuous control settings, where learning agents govern system actions, balancing the trade-off between reward maximization and constraint satisfaction remains a significant challenge. Policy optimization methods often exhibit instability near constraint boundaries, resulting in suboptimal training performance. To address this issue, we introduce a novel approach that integrates an adaptive incentive mechanism in addition to the reward structure to stay within the constraint bound before approaching the constraint boundary. Building on this insight, we propose Incrementally Penalized Proximal Policy Optimization (IP3O), a practical algorithm that enforces a progressively increasing penalty to stabilize training dynamics. Through empirical evaluation on benchmark environments, we demonstrate the efficacy of IP3O compared to the performance of state-of-the-art Safe RL algorithms. Furthermore, we provide theoretical guarantees by deriving a bound on the worst-case error of the optimality achieved by our algorithm.

cs.LG↗

arXiv2025

Tackling Uncertainties in Multi-Agent Reinforcement Learning through Integration of Agent Termination Dynamics

Multi-Agent Reinforcement Learning (MARL) has gained significant traction for solving complex real-world tasks, but the inherent stochasticity and uncertainty in these environments pose substantial challenges to efficient and robust policy learning. While Distributional Reinforcement Learning has been successfully applied in single-agent settings to address risk and uncertainty, its application in MARL is substantially limited. In this work, we propose a novel approach that integrates distributional learning with a safety-focused loss function to improve convergence in cooperative MARL tasks. Specifically, we introduce a Barrier Function based loss that leverages safety metrics, identified from inherent faults in the system, into the policy learning process. This additional loss term helps mitigate risks and encourages safer exploration during the early stages of training. We evaluate our method in the StarCraft II micromanagement benchmark, where our approach demonstrates improved convergence and outperforms state-of-the-art baselines in terms of both safety and task completion. Our results suggest that incorporating safety considerations can significantly enhance learning performance in complex, multi-agent environments.

cs.LG↗

arXiv2024

Learning-Enabled Adaptive Voltage Protection Against Load Alteration Attacks On Smart Grids

Smart grids are designed to efficiently handle variable power demands, especially for large loads, by real-time monitoring, distributed generation and distribution of electricity. However, the grid's distributed nature and the internet connectivity of large loads like Heating Ventilation, and Air Conditioning (HVAC) systems introduce vulnerabilities in the system that cyber-attackers can exploit, potentially leading to grid instability and blackouts. Traditional protection strategies, primarily designed to handle transmission line faults are often inadequate against such threats, emphasising the need for enhanced grid security. In this work, we propose a Deep Reinforcement Learning (DRL)-based protection system that learns to differentiate any stealthy load alterations from normal grid operations and adaptively adjusts activation thresholds of the protection schemes. We train this adaptive protection scheme against an optimal and stealthy load alteration attack model that manipulates the power demands of HVACs at the most unstable grid buses to induce blackouts. We theoretically prove that the adaptive protection system trained in this competitive game setting can effectively mitigate any stealthy load alteration-based attack. To corroborate this, we also demonstrate the method's success in several real-world grid scenarios by implementing it in a hardware-in-loop setup.

eess.SY↗

arXiv2024

Enhancing Attack Resilience in Real-Time Systems through Variable Control Task Sampling Rates

Cyber-physical systems (CPSs) in modern real-time applications integrate numerous control units linked through communication networks, each responsible for executing a mix of real-time safety-critical and non-critical tasks. To ensure predictable timing behaviour, most safety-critical tasks are scheduled with fixed sampling periods, which supports rigorous safety and performance analyses. However, this deterministic execution can be exploited by attackers to launch inference-based attacks on safety-critical tasks. This paper addresses the challenge of preventing such timing inference or schedule-based attacks by dynamically adjusting the execution rates of safety-critical tasks while maintaining their performance. We propose a novel schedule vulnerability analysis methodology, enabling runtime switching between valid schedules for various control task sampling rates. Leveraging this approach, we present the Multi-Rate Attack-Aware Randomized Scheduling (MAARS) framework for preemptive fixed-priority schedulers, designed to reduce the success rate of timing inference attacks on real-time systems. To our knowledge, this is the first method that combines attack-aware schedule randomization with preserved control and scheduling integrity. The framework's efficacy in attack prevention is evaluated on automotive benchmarks using a Hardware-in-the-Loop (HiL) setup.

eess.SY↗

arXiv2024

Smart Grid Security: A Verified Deep Reinforcement Learning Framework to Counter Cyber-Physical Attacks

The distributed nature of smart grids, combined with sophisticated sensors, control algorithms, and data collection facilities at Supervisory Control and Data Acquisition (SCADA) centers, makes them vulnerable to strategically crafted cyber-physical attacks. These malicious attacks can manipulate power demands using high-wattage Internet of Things (IoT) botnet devices, such as refrigerators and air conditioners, or introduce false values into transmission line power flow sensor readings. Consequently, grids experience blackouts and high power flow oscillations. Existing grid protection mechanisms, originally designed to tackle natural faults in transmission lines and generator outages, are ineffective against such intelligently crafted attacks. This is because grid operators overlook potential scenarios of cyber-physical attacks during their design phase. In this work, we propose a safe Deep Reinforcement Learning (DRL)-based framework for mitigating attacks on smart grids. The DRL agent effectively neutralizes cyber-physical attacks on grid surfaces by triggering appropriate sequences of existing protection schemes. The safety of the DRL agent is formally verified through a reachability analysis method. Additionally, our framework is designed for deployment on CUDA-enabled GPU systems, which enables faster execution of these protection sequences and their real-time validation. Our framework establishes a new set of protection rules for grid models, successfully thwarting existing cyber-physical attacks.

cs.CR↗

arXiv2024

Causality-Driven Reinforcement Learning for Joint Communication and Sensing

The next-generation wireless network, 6G and beyond, envisions to integrate communication and sensing to overcome interference, improve spectrum efficiency, and reduce hardware and power consumption. Massive Multiple-Input Multiple Output (mMIMO)-based Joint Communication and Sensing (JCAS) systems realize this integration for 6G applications such as autonomous driving, as it requires accurate environmental sensing and time-critical communication with neighboring vehicles. Reinforcement Learning (RL) is used for mMIMO antenna beamforming in the existing literature. However, the huge search space for actions associated with antenna beamforming causes the learning process for the RL agent to be inefficient due to high beam training overhead. The learning process does not consider the causal relationship between action space and the reward, and gives all actions equal importance. In this work, we explore a causally-aware RL agent which can intervene and discover causal relationships for mMIMO-based JCAS environments, during the training phase. We use a state dependent action dimension selection strategy to realize causal discovery for RL-based JCAS. Evaluation of the causally-aware RL framework in different JCAS scenarios shows the benefit of our proposed framework over baseline methods in terms of the beamforming gain.

cs.IT↗

arXiv2024

Probabilistic Interval Analysis of Unreliable Programs

Advancement of chip technology will make future computer chips faster. Power consumption of such chips shall also decrease. But this speed gain shall not come free of cost, there is going to be a trade-off between speed and efficiency, i.e accuracy of the computation. In order to achieve this extra speed we will simply have to let our computers make more mistakes in computations. Consequently, systems built with these type of chips will possess an innate unreliability lying within. Programs written for these systems will also have to incorporate this unreliability. Researchers have already started developing programming frameworks for unreliable architectures as such. In the present work, we use a restricted version of C-type languages to model the programs written for unreliable architectures. We propose a technique for statically analyzing codes written for these kind of architectures. Our technique, which primarily focuses on Interval/Range Analysis of this type of programs, uses the well established theory of abstract interpretation. While discussing unreliability of hardware, there comes scope of failure of the hardware components implicitly. There are two types of failure models, namely: 1) permanent failure model, where the hardware stops execution on failure and 2) transient failure model, where on failure, the hardware continues subsequent operations with wrong operand values. In this paper, we've only taken transient failure model into consideration. The goal of this analysis is to predict the probability with which a program variable assumes values from a given range at a given program point.

cs.PL↗

arXiv2023

A Learning Assisted Method for Uncovering Power Grid Generation and Distribution System Vulnerabilities

Intelligent attackers can suitably tamper sensor/actuator data at various Smart grid surfaces causing intentional power oscillations, which if left undetected, can lead to voltage disruptions. We develop a novel combination of formal methods and machine learning tools that learns power system dynamics with the objective of generating unsafe yet stealthy false data based attack sequences. We enable the grid with anomaly detectors in a generalized manner so that it is difficult for an attacker to remain undetected. Our methodology, when applied on an IEEE 14 bus power grid model, uncovers stealthy attack vectors even in presence of such detectors.

cs.CR↗

arXiv2023

Concealing CAN Message Sequences to Prevent Schedule-based Bus-off Attacks

This work focuses on eliminating timing-side channels in real-time safety-critical cyber-physical network protocols like Controller Area Networks (CAN). Automotive Electronic Control Units (ECUs) implement predictable scheduling decisions based on task level response time estimation. Such levels of determinism exposes timing information about task executions and therefore corresponding message transmissions via the network buses (that connect the ECUs and actuators). With proper analysis, such timing side channels can be utilized to launch several schedule-based attacks that can lead to eventual denial-of-service or man-in-the-middle-type attacks. To eliminate this determinism, we propose a novel schedule obfuscation strategy by skipping certain control task executions and related data transmissions along with random shifting of the victim task instance. While doing this, our strategy contemplates the performance of the control task as well by bounding the number of control execution skips. We analytically demonstrate how the attack success probability (ASP) is reduced under this proposed attack-aware skipping and randomization. We also demonstrate the efficacy and real-time applicability of our attack-aware schedule obfuscation strategy Hide-n-Seek by applying it to synthesized automotive task sets in a real-time Hardware-in-loop (HIL) setup.

cs.CR↗

arXiv2022

Exploring The Resilience of Control Execution Skips against False Data Injection Attacks

Modern Cyber-Physical Systems (CPSs) are often designed as networked, software-based controller implementations which have been found to be vulnerable to network-level and physical level attacks. A number of research works have proposed CPS-specific attack detection schemes as well as techniques for attack resilient controller design. However, such schemes also incur platform-level overheads. In this regard, some recent works have leveraged the use of skips in control execution to enhance the resilience of a CPS against false data injection (FDI) attacks. In this paper, we provide an analytical discussion on when and how skipping a control execution can improve the resilience of the system against FDI attacks while maintaining the control performance requirement. We also propose a methodology to synthesize such optimal control execution patterns. To the best of our knowledge, no previous work has provided any quantitative analysis about the trade-off between attack resilience and control performance for such aperiodic control execution. Finally, we evaluate the proposed method on several safety-critical CPS benchmarks.

eess.SY↗

arXiv2022

A CAD Framework for Simulation of Network Level Attack on Platoons

Recent developments in the smart mobility domain have transformed automobiles into networked transportation agents helping realize new age, large-scale intelligent transportation systems (ITS). The motivation behind such networked transportation is to improve road safety as well as traffic efficiency. In this setup, vehicles can share information about their speed and/or acceleration values among themselves and infrastructures can share traffic signal data with them. This enables the connected vehicles (CVs) to stay informed about their surroundings while moving. However, the inter-vehicle communication channels significantly broaden the attack surface. The inter-vehicle network enables an attacker to remotely launch attacks. An attacker can create collision as well as hamper performance by reducing the traffic efficiency. Thus, security vulnerabilities must be taken into consideration in the early phase of the development cycle of CVs. To the best of our knowledge, there exists no such automated simulation tool using which engineers can verify the performance of CV prototypes in the presence of an attacker. In this work, we present an automated tool flow that facilitates false data injection attack synthesis and simulation on customizable platoon structure and vehicle dynamics. This tool can be used to simulate as well as design and verify control-theoretic light-weight attack detection and mitigation algorithms for CVs.

cs.CR↗

arXiv2021

Co-designing Intelligent Control of Building HVACs and Microgrids

Building loads consume roughly 40% of the energy produced in developed countries, a significant part of which is invested towards building temperature-control infrastructure. Therein, renewable resource-based microgrids offer a greener and cheaper alternative. This communication explores the possible co-design of microgrid power dispatch and building HVAC (heating, ventilation and air conditioning system) actuations with the objective of effective temperature control under minimised operating cost. For this, we attempt control designs with various levels of abstractions based on information available about microgrid and HVAC system models using the Deep Reinforcement Learning (DRL) technique. We provide control architectures that consider model information ranging from completely determined system models to systems with fully unknown parameter settings and illustrate the advantages of DRL for the design prescriptions.

eess.SY↗

arXiv2021

An RL-Based Adaptive Detection Strategy to Secure Cyber-Physical Systems

Increased dependence on networked, software based control has escalated the vulnerabilities of Cyber Physical Systems (CPSs). Detection and monitoring components developed leveraging dynamical systems theory are often employed as lightweight security measures for protecting such safety critical CPSs against false data injection attacks. However, existing approaches do not correlate attack scenarios with parameters of detection systems. In the present work, we propose a Reinforcement Learning (RL) based framework which adaptively sets the parameters of such detectors based on experience learned from attack scenarios, maximizing detection rate and minimizing false alarms in the process while attempting performance preserving control actions.

cs.CR↗

arXiv2020

PySchedCL: Leveraging Concurrency in Heterogeneous Data-Parallel Systems

In the past decade, high performance compute capabilities exhibited by heterogeneous GPGPU platforms have led to the popularity of data parallel programming languages such as CUDA and OpenCL. Such languages, however, involve a steep learning curve as well as developing an extensive understanding of the underlying architecture of the compute devices in heterogeneous platforms. This has led to the emergence of several High Performance Computing frameworks which provide high-level abstractions for easing the development of data-parallel applications on heterogeneous platforms. However, the scheduling decisions undertaken by such frameworks only exploit coarse-grained concurrency in data parallel applications. In this paper, we propose PySchedCL, a framework which explores fine-grained concurrency aware scheduling decisions that harness the power of heterogeneous CPU/GPU architectures efficiently. %, a feature which is not provided by existing HPC frameworks. We showcase the efficacy of such scheduling mechanisms over existing coarse-grained dynamic scheduling schemes by conducting extensive experimental evaluations for a Machine Learning based inferencing application.

cs.DC↗

arXiv2020

Skip to Secure: Securing Cyber-physical Control Loops with Intentionally Skipped Executions

We consider the problem of provably securing a given control loop implementation in the presence of adversarial interventions on data exchange between plant and controller. Such interventions can be thwarted using continuously operating monitoring systems and also cryptographic techniques, both of which consume network and computational resources. We provide a principled approach for intentional skipping of control loop executions which may qualify as a useful control theoretic countermeasure against stealthy attacks which violate message integrity and authenticity. As is evident from our experiments, such a control theoretic counter-measure helps in lowering the cryptographic security measure overhead and resulting resource consumption in Control Area Network (CAN) based automotive CPS without compromising performance and safety.

cs.CR↗