Search

Deep Neural Networks (DNNs) are increasingly deployed across applications. However, ensuring their reliability remains a challenge, and in many situations, alternative models with similar functionality and accuracy are available. Traditional accuracy-based evaluations often fail to capture behavioral differences between models, especially with limited test datasets, making it difficult to select or combine models effectively. Differential testing addresses this by generating test inputs that expose discrepancies in DNN model behavior. However, existing approaches face significant limitations: many rely on model internals or are constrained by available seed inputs. To address these challenges, we propose DiffGAN, a black-box test image generation approach for differential testing of DNN models. DiffGAN leverages a Generative Adversarial Network (GAN) and the Non-dominated Sorting Genetic Algorithm II to generate diverse and valid triggering inputs that reveal behavioral discrepancies between models. DiffGAN employs two custom fitness functions, focusing on diversity and divergence, to guide the exploration of the GAN input space and identify discrepancies between models' outputs. By strategically searching this space, DiffGAN generates inputs with specific features that trigger differences in model behavior. DiffGAN is black-box, making it applicable in more situations. We evaluate DiffGAN on eight DNN model pairs trained on widely used image datasets. Our results show DiffGAN significantly outperforms a SOTA baseline, generating four times more triggering inputs, with greater diversity and validity, within the same budget. Additionally, the generated inputs improve the accuracy of a machine learning-based model selection mechanism, which selects the best-performing model based on input characteristics and can serve as a smart output voting mechanism when using alternative models.

Agile methodologies have gained significant traction in the software development industry, promising increased flexibility and responsiveness to changing requirements. However, their applicability to safety-critical systems, particularly in the automotive sector, remains a topic of debate. This paper examines the benefits and challenges of implementing agile methods in the automotive industry through a comprehensive review of relevant literature and case studies. Our findings highlight the potential advantages of agile approaches, such as improved collaboration and faster time-to-market, as well as the inherent challenges, including safety compliance and cultural resistance. By synthesizing existing research and practical insights, this paper aims to provide an understanding of the role of agile methods in shaping the future of automotive software development.

AUTOSAR (AUTomotive Open System ARchitecture) is an open industry standard for the automotive sector. It defines the three-layered automotive software architecture. One of these layers is the application layer, where functional behaviors are encapsulated in Software Components (SW-Cs). Inside SW-Cs, a set of runnable entities represents the internal behavior and is realized as a set of tasks. To address AUTOSAR's lack of support for modeling behaviors of runnables, languages such as Simulink are employed. Simulink simulations assume Simulink block behaviors are completed in zero execution time, while real execution requires a finite execution time. This timing mismatch can result in failures to detect unexpected runtime behaviors during the simulation phase. This paper extends the Simulink environment to model the timing properties of tasks. We present a Simulink block that can schedule tasks with non-zero simulation times. It enables a more realistic analysis during model development., Comment: 21 pages

Deep Reinforcement Learning (DRL) has made significant advancements in various fields, such as autonomous driving, healthcare, and robotics, by enabling agents to learn optimal policies through interactions with their environments. However, the application of DRL in safety-critical domains presents challenges, particularly concerning the safety of the learned policies. DRL agents, which are focused on maximizing rewards, may select unsafe actions, leading to safety violations. Runtime safety monitoring is thus essential to ensure the safe operation of these agents, especially in unpredictable and dynamic environments. This paper introduces SMARLA, a black-box safety monitoring approach specifically designed for DRL agents. SMARLA utilizes machine learning to predict safety violations by observing the agent's behavior during execution. The approach is based on Q-values, which reflect the expected reward for taking actions in specific states. SMARLA employs state abstraction to reduce the complexity of the state space, enhancing the predictive capabilities of the monitoring model. Such abstraction enables the early detection of unsafe states, allowing for the implementation of corrective and preventive measures before incidents occur. We quantitatively and qualitatively validated SMARLA on three well-known case studies widely used in DRL research. Empirical results reveal that SMARLA is accurate at predicting safety violations, with a low false positive rate, and can predict violations at an early stage, approximately halfway through the execution of the agent, before violations occur. We also discuss different decision criteria, based on confidence intervals of the predicted violation probabilities, to trigger safety mechanisms aiming at a trade-off between early detection and low false positive rates.

Deep Reinforcement Learning (DRL) algorithms have been increasingly employed during the last decade to solve various decision-making problems such as autonomous driving and robotics. However, these algorithms have faced great challenges when deployed in safety-critical environments since they often exhibit erroneous behaviors that can lead to potentially critical errors. One way to assess the safety of DRL agents is to test them to detect possible faults leading to critical failures during their execution. This raises the question of how we can efficiently test DRL policies to ensure their correctness and adherence to safety requirements. Most existing works on testing DRL agents use adversarial attacks that perturb states or actions of the agent. However, such attacks often lead to unrealistic states of the environment. Their main goal is to test the robustness of DRL agents rather than testing the compliance of agents' policies with respect to requirements. Due to the huge state space of DRL environments, the high cost of test execution, and the black-box nature of DRL algorithms, the exhaustive testing of DRL agents is impossible. In this paper, we propose a Search-based Testing Approach of Reinforcement Learning Agents (STARLA) to test the policy of a DRL agent by effectively searching for failing executions of the agent within a limited testing budget. We use machine learning models and a dedicated genetic algorithm to narrow the search towards faulty episodes. We apply STARLA on Deep-Q-Learning agents which are widely used as benchmarks and show that it significantly outperforms Random Testing by detecting more faults related to the agent's policy. We also investigate how to extract rules that characterize faulty episodes of the DRL agent using our search results. Such rules can be used to understand the conditions under which the agent fails and thus assess its deployment risks.

Deep Neural Networks (DNNs) have been extensively used in many areas including image processing, medical diagnostics, and autonomous driving. However, DNNs can exhibit erroneous behaviours that may lead to critical errors, especially when used in safety-critical systems. Inspired by testing techniques for traditional software systems, researchers have proposed neuron coverage criteria, as an analogy to source code coverage, to guide the testing of DNN models. Despite very active research on DNN coverage, several recent studies have questioned the usefulness of such criteria in guiding DNN testing. Further, from a practical standpoint, these criteria are white-box as they require access to the internals or training data of DNN models, which is in many contexts not feasible or convenient. In this paper, we investigate black-box input diversity metrics as an alternative to white-box coverage criteria. To this end, we first select and adapt three diversity metrics and study, in a controlled manner, their capacity to measure actual diversity in input sets. We then analyse their statistical association with fault detection using four datasets and five DNN models. We further compare diversity with state-of-the-art white-box coverage criteria. Our experiments show that relying on the diversity of image features embedded in test input sets is a more reliable indicator than coverage criteria to effectively guide the testing of DNNs. Indeed, we found that one of our selected black-box diversity metrics far outperforms existing coverage criteria in terms of fault-revealing capability and computational time. Results also confirm the suspicions that state-of-the-art coverage metrics are not adequate to guide the construction of test input sets to detect as many faults as possible with natural inputs.

Software Product Lines (SPLs) are families of related software products developed from a common set of artifacts. Most existing analysis tools can be applied to a single product at a time, but not to an entire SPL. Some tools have been redesigned/re-implemented to support the kind of variability exhibited in SPLs, but this usually takes a lot of effort, and is error-prone. Declarative analyses written in languages like Datalog have been collectively lifted to SPLs in prior work, which makes the process of applying an existing declarative analysis to a product line more straightforward. In this paper, we take an existing declarative analysis (behaviour alteration) written in the Grok declarative language, port it to Datalog, and apply it to a set of automotive software product lines from General Motors. We discuss the design of the analysis pipeline used in this process, present its scalability results, and provide a means to visualize the analysis results for a subset of products filtered by feature expression. We also reflect on some of the lessons learned throughout this project., Comment: MODELS'21 pre-print

The disposal of incinerator bottom ash usually produced by the incineration of inorganic constituents of the municipal solid wastes (MSW) in an eco-friendly way is one of the issues of concern, globally. By the way, the present study is related to the utilization of MSW incinerator bottom ash and recycled demolition waste aggregate as partial replacement materials for fine and coarse aggregate, respectively to produce eco-friendly concrete. This study adopted an innovative pretreatment technique for stabilizing the MSW incinerator bottom ash. Five distinct M20 grade concrete mixes were produced with different proportions of fine aggregate, MSW incinerator bottom ash, coarse aggregate, and recycled demolition waste aggregate along with cement and water. The incinerator bottom ash was replaced at 5% and 10% quantities with fine aggregate and the recycled demolition waste aggregate was replaced at 40% and 60% of the weight of the coarse aggregate. The strength and durability properties of the M20 grade concrete were analyzed. It was noticed that the strength and durability properties of the eco-friendly concrete specimens produced by incorporating 5% - incinerator bottom ash and 40% - recycled demolition waste aggregate were superior to that of the control mix concrete. Laboratory tank leaching tests showed that the eco-friendly concrete do not pose any significant environmental hazard. Furthermore, the microstructural analysis through scanning electron microscope (SEM) images, revealed dense aggregate paste matrix interfaces with less micro-pores and insignificant micro-cracks due to the incorporation of incinerator bottom ash as a partial replacement to the fine aggregate., Comment: Accepted for publication in Architecture, Structures and Construction Journal

In the literature, the definition of product in a Software Product Line (SPL) is based upon the notion of consistency of the constraints, imposed by variability and traceability relations on the elements of the SPL. In this paper, we contend that consistency does not model the natural semantics of the implementability relation between problem and solution spaces correctly. Therefore, we define when a feature can be {\em derived} from a set of components . Using this, we define a product of the SPL by a pair, where all the features in the specification are derived from the components in the architecture. This notion of derivability is formulated in a simple yet expressive, abstract model of a productline with traceability relation. We then define a set of SPL analysis problems and show that these problems can be encoded as Quantified Boolean Formulas. Then, QSAT solvers like QUBE can be used to solve the analysis problems. We illustrate the methodology on a small fragment of a realistic productline.

The Call admission control (CAC) is one of the Radio Resource Management (RRM) techniques plays instrumental role in ensuring the desired Quality of Service (QoS) to the users working on different applications which have diversified nature of QoS requirements. This paper proposes a fuzzy neural approach for call admission control in a multi class traffic based Next Generation Wireless Networks (NGWN). The proposed Fuzzy Neural Call Admission Control (FNCAC) scheme is an integrated CAC module that combines the linguistic control capabilities of the fuzzy logic controller and the learning capabilities of the neural networks .The model is based on Recurrent Radial Basis Function Networks (RRBFN) which have better learning and adaptability that can be used to develop the intelligent system to handle the incoming traffic in the heterogeneous network environment. The proposed FNCAC can achieve reduced call blocking probability keeping the resource utilisation at an optimal level. In the proposed algorithm we have considered three classes of traffic having different QoS requirements. We have considered the heterogeneous network environment which can effectively handle this traffic. The traffic classes taken for the study are Conversational traffic, Interactive traffic and back ground traffic which are with varied QoS parameters. The paper also presents the analytical model for the CAC .The paper compares the call blocking probabilities for all the three types of traffic in both the models. The simulation results indicate that compared to Fuzzy logic based CAC, Conventional CAC, The simulation results are optimistic and indicates that the proposed FNCAC algorithm performs better where the call blocking probability is minimal when compared to other two methods., Comment: Journal of Computing online at https://sites.google.com/site/journalofcomputing/

The Call admission control (CAC) is one of the Radio Resource Management (RRM) techniques that plays influential role in ensuring the desired Quality of Service (QoS) to the users and applications in next generation networks. This paper proposes a fuzzy neural approach for making the call admission control decision in multi class traffic based Next Generation Wireless Networks (NGWN). The proposed Fuzzy Neural call admission control (FNCAC) scheme is an integrated CAC module that combines the linguistic control capabilities of the fuzzy logic controller and the learning capabilities of the neural networks. The model is based on recurrent radial basis function networks which have better learning and adaptability that can be used to develop intelligent system to handle the incoming traffic in an heterogeneous network environment. The simulation results are optimistic and indicates that the proposed FNCAC algorithm performs better than the other two methods and the call blocking probability is minimal when compared to other two methods., Comment: International Journal of Computer Science Issues online at http://ijcsi.org/articles/A-QoS-Provisioning-Recurrent-Neural-Network-based-Call-Admission-Control-for-beyond-3G-Networks.php

The Next Generation Wireless Networks (NGWN) will be heterogeneous in nature where the different Radio Access Technologies (RATs) operate together .The mobile terminals operating in this heterogeneous environment will have different QoS requirements to be handled by the system. These QoS requirements are determined by a set of QoS parameters. The radio resource management is one of the key challenges in NGWN. Call admission control is one of the radio resource management technique plays instrumental role in ensure the desired QoS to the users working on different applications which have diversified QoS requirements from the wireless networks . The call blocking probability is one such QoS parameter for the wireless network. For better QoS it is desirable to reduce the call blocking probability. In this customary scenario it is highly desirable to obtain analytic Performance model. In this paper we propose a higher order Markov chain based performance model for call admission control in a heterogeneous wireless network environment. In the proposed algorithm we have considered three classes of traffic having different QoS requirements and we have considered the heterogeneous network environment which includes the RATs that can effectively handle applications like voice calls, Web browsing and file transfer applications which are with varied QoS parameters. The paper presents the call blocking probabilities for all the three types of traffic both for fixed and varied traffic scenario., Comment: 6 pages IEEE format, International Journal of Computer Science and Information Security, IJCSIS December 2009, ISSN 1947 5500, http://sites.google.com/site/ijcsis/

An interview with marketing professor S. Ramesh Kumar of Indian Institute of Management (IIM) in Bangalore, India is presented. When asked what are the effects of social media, he refers to the hedonism as the negative consequence and consumer education as the positive consequence. Kumar says that sociability and search concepts work together. He argues that the views of the company on social media will depend of their marketing strategy.