Search

Up to 10% of memory-safety vulnerabilities in languages like C and C++ stem from uninitialized variables. This work addresses the prevalence and lack of adequate software mitigations for uninitialized memory issues, proposing architectural protections in hardware. Capability-based addressing, such as the University of Cambridge's CHERI, mitigates many memory defects, including spatial and temporal safety violations at an architectural level. However, current CHERI designs do not handle undefined behavior from uninitialized variables. We extend the CHERI capability model to include "conditional capabilities", enabling memory-access policies based on prior operations. This allows enforcement of policies that satisfy memory safety objectives such as "no reads to memory without at least one prior write" (Write-before-Read). We present our architecture extension, compiler support, and a detailed evaluation of our approach using the QEMU full-system simulator and our modified FPGA-based CHERI-RISCV softcore. Our evaluation shows Write-before-Read conditional capabilities are practical, with high detection accuracy while adding a small (~3.5%) overhead to the existing CHERI architecture.

Rust is a popular memory-safe systems programming language. In order to interact with hardware or call into non-Rust libraries, Rust provides \emph{unsafe} language features that shift responsibility for ensuring memory safety to the developer. Failing to do so, may lead to memory safety violations in unsafe code which can violate safety of the entire application. In this work we explore in-process isolation with Memory Protection Keys as a mechanism to shield safe program sections from safety violations that may happen in unsafe sections. Our approach is easy to use and comprehensive as it prevents heap and stack-based violations. We further compare process-based and in-process isolation mechanisms and the necessary requirements for data serialization, communication, and context switching. Our results show that in-process isolation can be effective and efficient, permits for a high degree of automation, and also enables a notion of application rewinding where the safe program section may detect and safely handle violations in unsafe code.

Memory-related errors remain an important cause of software vulnerabilities. While mitigation techniques such as using memory-safe languages are promising solutions, these do not address software resilience and availability. In this paper, we propose a solution to build resilience against memory attacks into software, which contributes to environmental sustainability and security.

Until recently, conventional biochemical staining had the undisputed status as well-established benchmark for most biomedical problems related to clinical diagnostics, fundamental research and biotechnology. Despite this role as gold-standard, staining protocols face several challenges, such as a need for extensive, manual processing of samples, substantial time delays, altered tissue homeostasis, limited choice of contrast agents for a given sample, 2D imaging instead of 3D tomography and many more. Label-free optical technologies, on the other hand, do not rely on exogenous and artificial markers, by exploiting intrinsic optical contrast mechanisms, where the specificity is typically less obvious to the human observer. Over the past few years, digital staining has emerged as a promising concept to use modern deep learning for the translation from optical contrast to established biochemical contrast of actual stainings. In this review article, we provide an in-depth analysis of the current state-of-the-art in this field, suggest methods of good practice, identify pitfalls and challenges and postulate promising advances towards potential future implementations and applications., Comment: Review article, 4 main Figures, 3 Tables, 2 supplementary figures

Deep learning (DL) shows notable success in biomedical studies. However, most DL algorithms work as a black box, exclude biomedical experts, and need extensive data. We introduce the Self-Enhancing Multi-Photon Artificial Intelligence (SEMPAI), that integrates hypothesis-driven priors in a data-driven DL approach for research on multiphoton microscopy (MPM) of muscle fibers. SEMPAI utilizes meta-learning to optimize prior integration, data representation, and neural network architecture simultaneously. This allows hypothesis testing and provides interpretable feedback about the origin of biological information in MPM images. SEMPAI performs joint learning of several tasks to enable prediction for small datasets. The method is applied on an extensive multi-study dataset resulting in the largest joint analysis of pathologies and function for single muscle fibers. SEMPAI outperforms state-of-the-art biomarkers in six of seven predictive tasks, including those with scarce data. SEMPAI's DL models with integrated priors are superior to those without priors and to prior-only machine learning approaches.

Vehicle-to-Everything (V2X) communication is receiving growing attention from industry and academia as multiple pilot projects explore its capabilities and feasibility. With about 50\% of global road vehicle exports coming from the European Union (EU), and within the context of EU legislation around security and data protection, V2X initiatives must consider security and privacy aspects across the system stack, in addition to road safety. Contrary to this principle, our survey of relevant standards, research outputs, and EU pilot projects indicates otherwise; we identify multiple security and privacy related shortcomings and inconsistencies across the standards. We conduct a root cause analysis of the reasons and difficulties associated with these gaps, and categorize the identified security and privacy issues relative to these root causes. As a result, our comprehensive analysis sheds lights on a number of areas that require improvements in the standards, which are not explicitly identified in related work. Our analysis fills gaps left by other related surveys, which are focused on specific technical areas but not necessarily point out underlying root issues in standard specifications. We bring forward recommendations to address these gaps for the overall improvement of security and safety in vehicular communication.

Despite the increasing awareness from academia, civil society and media to the issue of child manipulation online, the current EU regulatory system fails at providing sufficient levels of protection. Given the universality of the issue, there is a need to combine and further these scattered efforts into a unitary, multidisciplinary theory of digital manipulation that identifies causes and effects, systematizes the technical and legal knowledge on manipulative and addictive tactics, and to find effective regulatory mechanisms to fill the legislative gaps. In this paper we discuss manipulative and exploitative strategies in the context of online games for children, suggest a number of possible reasons for the failure of the applicable regulatory system, propose an "upgrade" for the regulatory approach to address these risks from the perspective of freedom of thought, and present and discuss technological approaches that allow for the development of games that verifiably protect the privacy and freedoms of players.

This paper presents an approach to provide strong assurance of the secure execution of distributed event-driven applications on shared infrastructures, while relying on a small Trusted Computing Base. We build upon and extend security primitives provided by Trusted Execution Environments (TEEs) to guarantee authenticity and integrity properties of applications, and to secure control of input and output devices. More specifically, we guarantee that if an output is produced by the application, it was allowed to be produced by the application's source code based on an authentic trace of inputs. We present an integrated open-source framework to develop, deploy, and use such applications across heterogeneous TEEs. Beyond authenticity and integrity, our framework optionally provides confidentiality and a notion of availability, and facilitates software development at a high level of abstraction over the platform-specific TEE layer. We support event-driven programming to develop distributed enclave applications in Rust and C for heterogeneous TEE, including Intel SGX, ARM TrustZone and Sancus. In this article we discuss the workings of our approach, the extensions we made to the Sancus processor, and the integration of our development model with commercial TEEs. Our evaluation of security and performance aspects show that TEEs, together with our programming model, form a basis for powerful security architectures for dependable systems in domains such as Industrial Control Systems and the Internet of Things, illustrating our framework's unique suitability for a broad range of use cases which combine cloud processing, mobile and edge devices, and lightweight sensing and actuation., Comment: 41 pages main text + 4 pages appendix, first co-authorship between Gianluca Scopelliti and Sepideh Pouyanrad, source code available at https://github.com/AuthenticExecution

Security and safety are intertwined concepts in the world of computing. In recent years, the terms "sustainable security" and "sustainable safety" came into fashion and are being used referring to a variety of systems properties ranging from efficiency to profitability, and sometimes meaning that a product or service is good for people and planet. This leads to confusing perceptions of products where customers might expect a sustainable product to be developed without child labour, while the producer uses the term to signify that their new product uses marginally less power than the previous generation of that products. Even in research on sustainably safe and secure ICT, these different notions of terminology are prevalent. As researchers we often work towards optimising our subject of study towards one specific sustainability metric - let's say energy consumption - while being blissfully unaware of, e.g., social impacts, life-cycle impacts, or rebound effects of such optimisations. In this paper I dissect the idea of sustainable safety and security, starting from the questions of what we want to sustain, and for whom we want to sustain it. I believe that a general "people and planet" answer is inadequate here because this form of sustainability cannot be the property of a single industry sector but must be addressed by society as a whole. However, with sufficient understanding of life-cycle impacts we may very well be able to devise research and development efforts, and inform decision making processes towards the use of integrated safety and security solutions that help us to address societal challenges in the context of the climate and ecological crises, and that are aligned with concepts such as intersectionality and climate justice. Of course, these solutions can only be effective if they are embedded in societal and economic change towards more frugal uses of data and ICT.

Reliably detecting diseases using relevant biological information is crucial for real-world applicability of deep learning techniques in medical imaging. We debias deep learning models during training against unknown bias - without preprocessing/filtering the input beforehand or assuming specific knowledge about its distribution or precise nature in the dataset. We use control regions as surrogates that carry information regarding the bias, employ the classifier model to extract features, and suppress biased intermediate features with our custom, modular DecorreLayer. We evaluate our method on a dataset of 952 lung computed tomography scans by introducing simulated biases w.r.t. reconstruction kernel and noise level and propose including an adversarial test set in evaluations of bias reduction techniques. In a moderately sized model architecture, applying the proposed method to learn from data exhibiting a strong bias, it near-perfectly recovers the classification performance observed when training with corresponding unbiased data.

The use of unsafe programming languages still remains one of the major root causes of software vulnerabilities. Although well-known defenses that detect and mitigate memory-safety related issues exist, they don't address the challenge of software resilience, i.e., whether a system under attack can continue to carry out its function when subjected to malicious input. We propose secure rollback of isolated domains as an efficient and secure method of improving the resilience of software targeted by run-time attacks. We show the practicability of our methodology by realizing a software library for Secure Domain Rollback (SDRoB) and demonstrate how SDRoB can be applied to real-world software.

Modern vehicles are governed by a network of Electronic Control Units (ECUs), which are programmed to sense inputs from the driver and the environment, to process these inputs, and to control actuators that, e.g., regulate the engine or even control the steering system. ECUs within a vehicle communicate via automotive bus systems such as the Controller Area Network (CAN), and beyond the vehicles boundaries through upcoming vehicle-to-vehicle and vehicle-to-infrastructure channels. Approaches to manipulate the communication between ECUs for the purpose of security testing and reverse-engineering of vehicular functions have been presented in the past, all of which struggle with automating the detection of system change in response to message injection. In this paper we present our findings with fuzzing CAN networks, in particular while observing individual ECUs with a sensor harness. The harness detects physical responses, which we then use in a oracle functions to inform the fuzzing process. We systematically define fuzzers, fuzzing configurations and oracle functions for testing ECUs. We evaluate our approach based on case studies of commercial instrument clusters and with an experimental framework for CAN authentication. Our results show that the approach is capable of identifying interesting ECU states with a high level of automation. Our approach is applicable in distributed cyber-physical systems beyond automotive computing., Comment: Appeared in 2019 International Workshop on Attacks and Defenses for Internet-of-Things (ADIoT) / International Workshop on the Secure Internet of Things (SIoT)

Clinical decision support using deep neural networks has become a topic of steadily growing interest. While recent work has repeatedly demonstrated that deep learning offers major advantages for medical image classification over traditional methods, clinicians are often hesitant to adopt the technology because its underlying decision-making process is considered to be intransparent and difficult to comprehend. In recent years, this has been addressed by a variety of approaches that have successfully contributed to providing deeper insight. Most notably, additive feature attribution methods are able to propagate decisions back into the input space by creating a saliency map which allows the practitioner to "see what the network sees." However, the quality of the generated maps can become poor and the images noisy if only limited data is available - a typical scenario in clinical contexts. We propose a novel decision explanation scheme based on CycleGAN activation maximization which generates high-quality visualizations of classifier decisions even in smaller data sets. We conducted a user study in which we evaluated our method on the LIDC dataset for lung lesion malignancy classification, the BreastMNIST dataset for ultrasound image breast cancer detection, as well as two subsets of the CIFAR-10 dataset for RBG image object recognition. Within this user study, our method clearly outperformed existing approaches on the medical imaging datasets and ranked second in the natural image setting. With our approach we make a significant contribution towards a better understanding of clinical decision support systems based on deep neural networks and thus aim to foster overall clinical acceptance., Comment: 49 pages, 13 figures, 18 pages appendix, accepted manuscript

Computer systems often provide hardware support for isolation mechanisms like privilege levels, virtual memory, or enclaved execution. Over the past years, several successful software-based side-channel attacks have been developed that break, or at least significantly weaken the isolation that these mechanisms offer. Extending a processor with new architectural or micro-architectural features, brings a risk of introducing new such side-channel attacks. This paper studies the problem of extending a processor with new features without weakening the security of the isolation mechanisms that the processor offers. We propose to use full abstraction as a formal criterion for the security of a processor extension, and we instantiate that criterion to the concrete case of extending a microprocessor that supports enclaved execution with secure interruptibility of these enclaves. This is a very relevant instantiation as several recent papers have shown that interruptibility of enclaves leads to a variety of software-based side-channel attacks. We propose a design for interruptible enclaves, and prove that it satisfies our security criterion. We also implement the design on an open-source enclave-enabled microprocessor, and evaluate the cost of our design in terms of performance and hardware size., Comment: Extended version of the paper "Provably Secure Isolation for Interruptible Enclaved Execution on Small Microprocessors"