As autonomous vehicles gain popularity, ensuring their safety and security becomes important. Multi-agent Collaborative Perception (MACP) enhances autonomous cars' perception capabilities. Despite its benefits, MACP is vulnera-ble to adversarial attacks, wherein malicious agents introduce adversarial examples to compromise the perception of recipient vehicles. This paper investigates the vulnerability of MACP to adversarial attacks, analyzing both white-box (where the attacker has access to the target model's gradients) and black-box scenarios (where the attacker lacks such access). We introduce an Agent-Aware Defense System designed to minimize these vulnerabilities in generic scenario. Through the application of the Fast Gradient Signed Method and Transferability techniques, we generate perturbations to assess the robustness of MACP models. Our results indicate that white-box attacks are more effective than black-box attacks and that early-stage collaborative methods are particularly vulnerable to adversarial manipulation. This study shows that integrating robust defense mechanisms in the development of collaborative perception frameworks for autonomous vehicles is important.
The general goal of traffic modeling is to predict the size of traffic flows in individual roads of the modeled road network. A commonly employed technique is the user-equilibrium (UE) traffic assignment (TA). There are multiple algorithms for solving TA with different convergence speeds. One of the fastest algorithms is the Improved Traffic Assignment by Paired Alternative Segments (iTAPAS) algorithm. In this paper, an optimized implementation of the iTAP AS algorithm is described and compared to its original implementation and to the sequential and parallel versions of the B algorithm, which we implemented during our previous research. An important part of the paper is the discussion of the parallelization of the iTAP AS algorithm, which is our ultimate goal.
This paper presents a simulation-based comparison of deep learning and reinforcement learning (RL) architectures for automatic modulation classification (AMC) under low-SNR conditions and real-time constraints. We benchmark a robust convolutional neural network (CNN), propose a lightweight dilated convolutional network (DCN-MODNET), and evaluate an exploratory RL agent using a custom Gym simulation. Our goal is to assess trade-offs between accuracy, inference latency, and computational cost for deployment in cognitive radio networks (CRNs). Results show that the DCN-MODNET offers strong performance with reduced complexity, while the RL agent shows niche utility in ultra-low SNRs but remains unsuitable for real-time use.
Artificial intelligence is transforming science into a highly automated process, accelerating the pace of scientific discovery. We argue that automatically adapting simulation models to new data, knowledge about mechanisms, or research questions plays a key role in this change of paradigm. To this end, we explore the use of the MAPE- K (Monitor, Analyze, Plan, and Execute, over a shared Knowledge base) framework from self-adaptive software systems to realize a self-adaptive simulation model. Such a model will preserve its currency, reveal new insights, and provide enhanced predictions as well as effective feedback to the observed system. As a case study, we use a cell-biological model of glucose-stimulated insulin secretion. We conclude with a discussion relating our approach to digital twins and current developments in inferring simulation models from data that opens up questions to be pursued in future research.
Predicting how multiple road users move is crucial for autonomous driving, but accuracy, reliability, and clarity are often hard to balance. This paper reviews hybrid methods that combine kinematic, map, and behavioral information with data-driven models. It organizes hybrid methods by where constraints are applied and compares them with classical and deep models using standard datasets and metrics in a compact table. It then discusses three key deployment challenges, which are handling domain shifts, calibrating uncertainty and interpretability, and mitigating closed-loop degradation-and outline practical checks for integrating hybrids into planning pipelines. Its goal is to provide a concise, practice-oriented perspective that shows when hybrids improve reliability and where they add complexity, to guide the design of accurate and trustworthy trajectory predictors.
Crowdsensing platforms face a fundamental trade-off between data utility and participant privacy, where traditional approaches require users to expose sensitive identity information, creating barriers to widespread adoption. This paper presents a blockchain- based protocol that addresses these privacy concerns through pseudonymous participation and secure data exchange mechanisms. Our approach exploits smart contracts as trusted intermediaries to eliminate di-rect communication between data initiators and contributors, while employing asymmetric cryptography to enable secure key exchange without pre-established channels. The protocol enhances privacy through campaign-specific key pairs that remain unlinkable to participants' persistent identities, contex-tual separation of cryptographic identities, and anonymization sets that obscure individual actions within larger groups. We developed a prototype implementation to verify the correctness of the protocol and to evaluate gas consumption. Through simulation, we assess system performance under varying user dynamics and activity rates. Results confirm the viability of the proposal.
Real-time animal detection in search-and-rescue (SAR) operations represents a critical challenge for situational awareness systems, particularly when deploying lightweight solutions on resource-constrained edge computing platforms. Current detection methods suffer from computational bottlenecks that compromise either accuracy or real-time performance, limiting their effectiveness in time-critical rescue scenarios. This paper presents SVR-YOLO11, an enhanced detection network specifically optimized for real-time animal identification in SAR applications. The proposed architecture introduces three key in-novations: (1) a Slim-Neck design paradigm that preserves inter-channel connections while reducing computational complexity, (2) integration of Large Separable Kernel Attention (LSKA) modules with dynamic upsampling to improve detection accuracy for morphologically similar animal species across diverse natural backgrounds, and (3) direct deployment capability on edge devices without requiring complex model conversion processes. Experimental validation on the Animals. v4i dataset demonstrates that SVR-YOLO11 achieves superior performance with 68.3% mAP50-95 while maintaining only 2.843M parameters and 6.5 GFLOPs computational cost. Real-time testing on multiple edge computing platforms, including Raspberry Pi and NVIDIA Jetson devices, confirms the network's practical applicability for field deployment. The system's effectiveness is further validated through a comprehensive head-mounted display demonstration using Unity Engine simulations that replicate dynamic SAR environments. These results establish SVR-YOLO11 as a ro-bust solution for real-time animal detection in distributed edge computing scenarios, directly supporting enhanced situational awareness in critical rescue operations.
This paper proposes a method for formally modeling and analyzing mutual exclusion algorithms. The process starts with Uppaal timed automata and model checking. A Uppaal model is then reduced to a Java actor program to overcome the scalability issues associated with exhaustive verification. The method is applied to mutual exclusion in anonymous shared memory (ASM). ASM is challenging because concurrent/parallel processes do not agree on the names of the shared registers. Rather, each process uses a permutation of the register names. Processes have unknown identifiers and are symmetric, that is, they execute the same code. ASM is natural in bio-inspired, molecular applications. As a case study, the paper investigates a recent ASM mutual exclusion solution proposed by Taubenfeld. The algorithm proves correct for N=2 processes when m≥7 (m odd) registers are used. The algorithm does not extend to N≥3 processes. As a new contribution, this paper develops and formalizes a solution (TT - T2) that embeds Taubenfeld's algorithm for 2 processes (T2) in a tournament binary tree (TT) organization, where the basic algorithm acts as the arbitration unit in TT nodes. The paper demonstrates, through simulations of the reduced and scalable actor program, that the TT - T2 represents an effective solution for mutual exclusion in ASM for N≥2 processes.
Forest fires are destructive events with significant social and economic consequences. Simulation may provide an invaluable tool for improving the effectiveness of fire suppression strategies. Cellular automata (CA) have been used for modeling fire behavior providing a discretization of space and enabling a local description of fire within each CA cell. In this paper we introduce the πHYFLOW formalism to model forest fires using CA. This formalism supports the process interaction worldview enabling the definition of dynamic processes/agents that can be created and destroyed at runtime. Simulation results demonstrate that the number of active fire agents is only a small fraction of the total number of CA cells, highlighting the efficiency of our method in representing large-scale forest fire scenarios.
Modern simulation environments require advanced tools for real-time data visualization and interactive analysis. This paper presents LiveOppServation, an open-source Dash-based frontend for OMNeT++ simulations that enables continuous runtime monitoring and visualization of simulation data. Our system addresses limitations in traditional post-processing approaches by providing three key capabilities: real-time evaluation of vector data with timestamps, automatic generation of live visualizations, and conditional data-based breakpoints for targeted event analysis. Built on gRPC and Protocol Buffers, the architecture supports scalable many-to-many mapping between simulations and frontends, allowing simultaneous monitoring of multiple simulations or collaborative analysis of a single simulation from different perspectives. The design emphasizes flexibility through modular components and extensibility for integration with other simulation platforms.
Biochemical systems are characterized by rich stochastic behaviors and complex interaction patterns, often modeled through rule-based frameworks such as BioNetGen. Despite the expressiveness of such high-level specifications, simulating large-scale models remains computationally prohibitive. In this paper, we present a model-driven framework for the parallel and distributed execution of stochastic chemical reaction networks, enabled by a model-to-model transformation of BioNetGen descriptions. The transformation produces an intermediate representation based on the Actor Model, in which actors encapsulate local state and asynchronous communication, aligning with the concurrency of biochemical processes. From this representation, we generate executable code targeting ROOT-Sim, a speculative parallel discrete-event simulation (PDES) environment based on Time Warp. We propose an exact parallel implementation of the Stochastic Simulation Algorithm (SSA), employing reaction partitioning to minimize inter-process dependencies and leveraging an event-exchange protocol to ensure consistency and atomicity of reactant consumption across logical processes. We introduce refined rollback and reaction rescheduling mechanisms to address potential correctness issues such as reactant overconsumption. Extensive experiments on well-established models, such as FcεRI, demonstrate substantial speedups over sequential methods.
This paper introduces a formal fragmentation of the RAMSAS methodology aimed at supporting modular, scalable, and automated reliability analysis of complex systems through simulation. The proposed approach contributes in three main di-rections: (i) a formal specification of RAMSAS method fragments within the Method Engineering paradigm, (ii) the definition of a dedicated algebra to govern their composition, and (iii) the development of automated fault analysis mechanisms driven by simulation results. By enabling systematic decomposition and controlled recomposition, the framework extends RAMSAS with enhanced composability, support for distributed execution of simulations, and automatic generation of Fault Trees (FT). A dedicated metamodel together with algebraic operators provides a foundation for formal reasoning on method fragments and their seamless integration into model-based simulation workflows. The approach has been specified and is currently being evaluated within the “Space It Up” project, funded by the Italian Space Agency (ASI) and the Italian Ministry of University and Research (MUR).
In a ridesharing service, driver-rider matching affects the business in several ways. Not only does it cost money to find a route on a map, but it also has an impact on the business in terms of driver income and rider waiting times, and therefore their satisfaction. It is not practical to field test the matching. For example, an A/B test would take days. A simulator greatly accelerates matching experiments. We modeled ridesharing and developed a simulator based on it. The event-driven simulator enabled us to test matching algorithms with a variety of parameters while utilizing routing systems, such as OSRM. Tested matching techniques are back-to-back and reassignment, that determine the matching targets. The simulations showed that back-to-back reduced the average waiting time for riders by up to around 40% and reassignment reduced it by up to around 15%. It is also the contribution of this paper to show what ridesharing operators are thinking about in order to technically optimize their business.
Digital twin technology is revolutionizing cyber-security by providing real-time, data-driven replicas of ICT infrastructures without impacting live production systems. We present NotLine, a non-intrusive, fully automated pipeline platform that builds and updates digital twins through the continuous passive collection of multiprotocol network traffic metadata. NotLine filters and normalizes the data to remove noise and then correlates events to generate a dynamic topology graph. This non-intrusive approach enhances network monitoring and mitigates the risks and overhead associated with active scanning. It also offers superior scalability and enables continuous threat hunting, risk assessment, and accelerated remediation. The resulting digital twin extends traditional static inventories with real-time vulnerability mapping through CVE lookups. It also supports AI -driven adversary simulations based on Monte Carlo methods. Furthermore, we explore how integrating non-intrusive host-level telemetry, threat intelligence feeds, and re-inforcement learning can evolve the digital twin into a self-optimizing cybersecurity guardian. Experiments in production environments demonstrate that passive monitoring over extended periods, spanning multiple days, is essential for accurately capturing daily and weekly usage patterns across diverse protocol families. A quantitative analysis is presented that establishes benchmarks for digital twin fidelity in networked environments.
Modeling and simulation (M&S) have become in-dispensable “tools for thinking” across science, engineering, and politics, to explore, understand, and analyze complex systems. Increases in model size, model complexity, and simulation speed result in the growing need to channel users' cognitive focus toward the most creative and insightful M&S tasks. This paper proposes a conceptual model that aligns the core M&S stages-model creation, simulation, and analysis-with Peirce's retroduction cycle, an integrated reasoning process combining abduction, deduction, and induction. We clarify the distinct cognitive strengths of each reasoning mode-creativity, soundness, reliability, generalizability, and computability-and systematically identify how computational and interactive interventions can support these cognitive activities. Our structured analysis highlights opportunities for hybrid reasoning systems, guiding users' cognitive focus strategically to enhance insight generation and creative problem-solving within complex M&S environments.
Computer simulation can be a useful tool for road traffic management and analysis. In many cases, it is necessary or at least desirable for the simulation to run as fast as possible. A common way how to speed up the simulation is to use a distributed or parallel computing environment, where combined power of multiple computers and/or processor cores is utilized. In that case, road network division is usually employed to divide the simulated road network into sub-networks, whose simulations are then performed by individual processes and/or threads of the distributed or parallel simulation. The quality of the road network division can significantly affect the performance of the resulting distributed or parallel simulation. There are dozens of division methods employing various algorithms, but mutual comparison of their performances is hindered by heterogeneity of their implementations and of the tests described in corresponding scientific papers. The methods are written in various programming languages, optimize various parameters of the road network division, and divide various road networks. In this paper, this issue is addressed by designing a tool for comparison of multiple division methods in a unified environment. The functioning of this ROad Network DIvision BEnchmark Tool (RONDIBET) is demonstrated on a case study of comparison of two third-party division methods.
In this study, to reduce the harm of the social and economic impacts of typhoons, a novel CNN-Transformer En-coder hybrid model is proposed to predict the typhoon trajectory for the previous preparation of typhoons, designed to support real-time simulation systems and improve preparedness against typhoon impacts. It combines the spatial feature extraction ability of a Convolutional Neural Network (CNN) with the temporal dependency modeling ability of a Transformer Encoder. Two different datasets, including the CMA tropical cyclone best track dataset and the ERA5 dataset, are integrated to improve the prediction accuracy and the diversity of data sources. Further-more, by incorporating critical environmental factors such as sea pressure and sea surface temperature, the predictive model can be integrated into real-time typhoon simulation systems to en-hance their dynamic accuracy and operational applicability. The model is validated by performance metrics such as mean square error, root mean squared error, coefficient of determination and case studies, demonstrating its adaptability to incomplete data and its ability to predict accurate trajectories. Subsequently, a comparative analysis between the CNN- Transformer model and the CNN-LSTM model demonstrates the better performance and enhanced predictive capabilities of the proposed approach.
This study leverages the Simulation of Urban MO-bility (SUMO) platform to conduct a comprehensive evaluation of vehicular gas emissions and fuel consumption across various urban traffic scenarios. Utilizing the OpenStreetMap (OSM) Web Wizard for scenario generation, a range of traffic configurations was modeled to reflect diverse urban environments. These config-urations incorporated multiple vehicle types-such as passenger cars, heavy-duty trucks, and public transportation-as well as a spectrum of driving behaviors. The research specifically examines the impact of using default or “quick-start” simulation settings provided by SUMO's OSM Web Wizard, assessing how these simplified configurations affect the accuracy of environmental and performance-related metrics. Findings indicate that while these default settings offer convenience and ease of use, they can lead to substantial deviations in emission and fuel consumption estimates when compared to more customized and calibrated simulations. This underscores the importance of careful scenario tuning and gives an important indication of how vehicle types and human driving behavior impact CO2 emission and fuel consumption.
Real-time embedded systems, fundamental to cyber-physical systems (CPS), employ microcontrollers, memory, and I/O devices to perform complex jobs within stringent time limits. A crucial component is the scheduler, which manages resources efficiently, ensuring timely job execution based on priority. Multiprocessor platforms, increasingly adopted in CPS, enhance performance, scalability, and fault tolerance by enabling parallel job execution and redundancy. Minimizing energy consumption is essential for extending battery life, especially in mobile CPS applications. However, most fault-tolerant strategies focus on either homogeneous multiprocessors or systems with only two processor types. This study introduces a heuristic scheduling technique, ECO-FAST, that addresses fault tolerance and energy management in systems with diverse processor types. Benchmark tests show that ECO-FAST outperforms a state-of-the-art approach by generating more secondary job copies to mitigate transient faults while significantly reducing energy consumption.
The recent disruptions in manufacturing value creation showed the fragility of currently dominating linear value chains. Transformation of manufacturing to a service-oriented approach is a promising solution to achieve resiliency and flexibility. Service-oriented production is enabled by the “Manufacturing Service” approach, which is a core element of the Manufacturing-as-a-Service (MaaS) paradigm. In aMaaS enabled manufacturing ecosystem, consumers can avail manu-facturing services from providers, making the products transit across networked MaaS providers. Hence, an efficient decision support system is required for MaaS consumers, to ensure the viability of production in terms of cost, technology, and further requirements such as sustainability prior to ordering aMaaS from providers. This paper proposes a platform architecture for a MaaS enabled manufacturing ecosystem. Essential elements in the platform and a methodology to leverage simulation models backed by ontologies for automated quantified discovery of MaaS from providers are introduced. We anticipate that such a decision support system will boost the adoption of the MaaS paradigm. A case study is used to outline the applicability of our solution in a distributed manufacturing ecosystem.
