This thesis explores the security of intermittent devices, embedded systems designed to retain their state across periods of power loss, for cases both when the device has an excess of available energy and when power loss is unavoidable. Existing work with intermittent systems has focused on the problems inherent to the intermittent paradigm and ignored the security implications of persistent state across periods of power loss. The security of these devices is closely linked to their unique operational characteristics and are addressed here in two studies. First, the presence of an energy harvester creates an opportunity to use excess energy, available when additional energy is harvested after the local energy reservoir is filled, to precompute security related operations. Precomputation powered by this excess energy can reduce the cost of expensive tasks during periods of energy scarcity, potentially enabling the use of expensive security operations on traditionally unsecured devices. Second, when energy is limited and intermittent operation is required, the secure storage of checkpoints is a necessity to protect against adversary manipulation of the system state. To examine the secure storage of checkpoints a protocol is implemented to ensure the integrity and authenticity of a device's checkpoints, and evaluated for its energy overhead and performance. The cost of properly ensuring the integrity and authenticity of these checkpoints is examined to identify the overhead necessary to execute intermittent operations in a secure manner. Taken together, these studies lay the groundwork for a comprehensive view of the current state of intermittent device security. / Master of Science

International Telemetering Conference Proceedings / October 18-21, 2004 / Town & Country Resort, San Diego, California / The embedded system related hardware technology has experienced rapid development, and it provided the software technology with a huge space for growth. Therefore using software approaches to perform GPS receiver functions in a powerful and generic hardware platform is becoming more feasible. In this paper, the software GPS receiver technology and the design basics of the software receiver are discussed. Further in the Matlab simulation environment, the implementation of a software receiver for replacing the processing functions of ASIC in traditional GPS receivers, i.e. RF front end and multi-channel correlator, is presented. Some simulation results and implementation details are included.

International Telemetering Conference Proceedings / October 21, 2002 / Town & Country Hotel and Conference Center, San Diego, California / Because of an ever-increasing need for performance and high predictability in modern real-time telemetry systems, the Java programming language is typically not considered a viable option for embedded software development. Nevertheless, the Java platform provides many features that can easily be applied to embedded telemetry systems that other development platforms cannot match. But obviously, there are pitfalls to be aware of. This paper will present an alternative solution to address today’s problems in real-time telemetry systems and will cover the following topics: • Java development platforms for the embedded world • Impact on software portability and reusability • Performance and optimization techniques • Direct access to hardware devices • Memory management and garbage collection • Network-centric component-oriented architecture • Real-time examples from past experience • Future developments

This report describes an embedded system designed to support the development of embedded augmented reality applications. It includes an integrated camera and built-in graphics acceleration hardware. An example augmented reality application serves as a demonstration of how these features are accessed, as well as providing an indication of the performance of the device. The embedded augmented reality development platform consists of the Gumstix Overo computer-on-module paired with the custom-built Overocam camera board. This device offers an ARM Cortex-A8 CPU running at 600 MHZ and 256 MB of RAM, along with the ability to capture VGA video at 30 frames per second. The device runs an operating system based on version 2.6.33 of the Linux kernel. The main feature of the device is the OMAP3530 multimedia applications processor from Texas Instruments. In addition to the ARM CPU, it provides an on-board 2D/3D graphics accelerator and a digital signal processor. It also includes a built-in camera peripheral interface, reducing the complexity of the camera board design. A working example of an augmented reality application is included as a demonstration of the device's capabilities. The application was designed to represent a basic augmented reality task: tracking a single marker and rendering a simple virtual object. It runs at around 8 frames per second when a marker is visible and 13 frames per second otherwise. The result of the project is a self-contained computing platform for vision-based augmented reality. It may either be used as-is or customised with additional hardware peripherals, depending on the requirements of the developer.

No description available.

Recent development in the field of distributed and decentralized cyber-physical systems led to emerge of DEECo model. As many DEECo use cases are embedded applications it is interesting to evaluate DEECo on embedded hardware. Currently there is only reference DEECo implementation which is written in Java thus cannot be used for embedded applications. As part of this thesis C++ DEECo mapping and embedded CDEECo++ framework were designed using FreeRTOS operating system for task scheduling and synchronization. An example application designed for the STM32F4 board demonstrates usability of the framework. This thesis contains description of the DEECo mapping into the C++ language, source codes of the CDEECo++ framework, documentation and example application including basic measurement of its real- time properties. Powered by TCPDF (www.tcpdf.org)

Over the past few years, the interest of remote wireless sensor networks has increased with the growth of Internet of Things technology. The wireless sensor network applications vary from tracking animal movement to controlling small electrical devices. Wireless sensors deployed in remote areas where the grid is unavailable are normally powered by batteries, inducing a limited lifespan for the sensor. This thesis work presents a solution to implement solar energy harvesting to a wireless sensor network. By gathering energy from the environment and using it in conjunction with an energy storage, the lifetime of a sensor node can be extended while at the same time reducing maintenance costs. To make sensor nodes in a network energy efficient, an adaptive controller of the nodes energy consumption can be used. A network consisting of a client node and a server node was created. The client node was powered by a small solar cell in conjunction with a capacitor. A linear-quadratic tracking algorithm was implemented to adaptively change the transmission rate for a node based on its current and previous battery level and the energy harvesting model. The implementation was done using only integers. To evaluate the system for extended run-times, the battery level was simulated using MATLAB. The system was simulated for different weather conditions. The simulation results show that the system is viable for both cloudy and sunny weather conditions. The integer linear-quadratic algorithm responds to change very abruptly in comparison to a floating point-version.

Real-time embedded systems can be found in a large number of devices we use, including safety-critical systems. Useful for their small size and low power consumption, they are also harder to protect against state-of-the-art attacks than general purpose systems due to their lack of hardware features. Even current defenses may not be applicable since instrumentation added to defend real-time embedded systems may cause them to miss their deadlines, rending them inoperable. We show that the static properties obtained by the scheduling policies can be used as security guarantees for the tasks composing the program. By completely securing a subset of the tasks of the program only using the scheduler policy, we remove the need to add external instrumentation on these tasks, reducing the amount of extra instructions needed to entirely protect the system. With less instrumentation, the overhead added by the defenses is reduced and can therefore be applied to a larger number of systems.

AUTOSAR (AUTomotive Open System Architecture) est un standard industriel mondial créé en 2003 dans le but de standardiser le développement des architectures logicielles automobiles. Il fournit un ensemble de concepts et définit une méthodologie commune pour le développement des logiciels embarqués automobiles. Les principales caractéristiques de ce standard sont la modularité et la « configurabilité» de logiciels qui permettent la réutilisation fonctionnelle des modules logiciels fournis par des fournisseurs différents. Cependant,le développement d’une application embarquée AUTOSAR nécessite la configuration d’un grand nombre de paramètres liés principalement au grand nombre de composants logiciels (software component« SWC ») de l’application. Cette configuration commence par l’étape d’allocation des SWCs à la plateforme matérielle (calculateursconnectés par des réseaux), jusqu’à l’étape de configuration de chaque calculateur et du réseau de communication. Différentes alternatives sont possibles pendant ces étapes de configuration etc chaque décision de conception peut impacter les performances temporelles du système, d’où la nécessité d’automatiser ces étapes de configuration et de développer un outil d’évaluation d’architectures.Dans ce travail de thèse, nous introduisons une approche holistique d’optimisation afin de synthétiser l’architecture E/E d’un système embarqué AUTOSAR. Cette approche se base sur des méthodes métaheuristique et heuristique. La méthode métaheuristique (i.e. algorithme génétique) a le rôle de trouver les allocations les plus satisfaisantes des SWCs aux calculateurs. A chaque allocation proposée, deux méthodes heuristiques sont développées afin de résoudre le problème de la configuration des calculateurs (le nombre de tâches et ses priorités, allocation des runnables aux tâches, etc.) et des réseaux de communication (le nombre de messages et ses priorités, allocation des « data-elements » aux messages,etc.). Afin d’évaluer les performances de chaque allocation, nous proposons une nouvelle méthode d’analyse pour calculer le temps de réponse des tâches, des runnables, et de bout-en-bout de tâches/runnables. L’approche d’exploration architecturale proposée par cette thèse considère le modèle des applications périodiques et elle est évaluée à l’aide d’applications génériques et industrielles. / AUTOSAR (AUTomotive Open System ARchitecture) has been created by automotive manufacturers, suppliers and tools developers in order to establish an open industry standard for automotive E/E(Electrical/Electronic) architectures. AUTOSAR provides a set of concepts and defines a common methodology to develop automotive software platforms. The key features of this standard are modularity and configurability of automotive software; this allows functional reuse of software modules provided by different suppliers and guarantees interoperability of these modules through standardized interfaces. However, the development of an embedded application according to AUTOSAR necessitates configuring a lot of parameters related to the large number of Software Components (SWCs), their allocations to the hardware platform and then, the configurationof each Electronic Control Unit (ECU). Different alternatives are possible during the design of such systems. Each implementation decision may impact system performance and needs therefore to be evaluated and compared against performance constraints and optimization goals. In this thesis, we introduce a holistic optimization approach to synthesizearchitecture E/E of an embedded AUTOSAR system. This approach is based on heuristic and metaheuristic methods. The metaheuristics (e.g. genetic algorithm) has the role to find the most satisfactory allocations of SWCs to ECUs. Each allocation step, two heuristics are developed to solve the problem of the ECU configuration (the number of tasks and priorities, allocation of runnables to tasks, etc.) and networks configuration (the number of messagesand priorities, allocation of data-elements to messages, etc.). In order to evaluate the performance of each allocation, we propose a new analysis method to calculate the response time of tasks, runnables, and end-to-end paths. The architectural exploration approach proposed by this thesis considers the model for periodic applications and is evaluated using generic and industrial applications.

No description available.