Body Area Sensor Networks (BANs) utilize wireless communicating sensor nodes attached to a human body for convenience, safety, and health applications.

Physiological characteristics of the body, such as the heart rate or ECG signals, are promising means to simplify the setup process and to improve security of BANs.

This thesis describes the design and implementation steps required to realize an ECG-based authentication protocol to identify sensor nodes attached to the same human body.

This thesis starts by reviewing the latest literature in the filed of Body Area Network, and looks at possible applications, challenges, and overall architecture of BANs.

It also addresses the design of a body-area sensor system, including the hardware setup, analogue and digital signal processing, and required ECG feature detection techniques.

A model-based design flow is applied, and strengths and limitations of each design step are discussed.

Real-world measured data originating from the implemented sensor system then are used to set up and parametrize a novel physiological authentication protocol for BANs.

The authentication protocol utilizes statistical properties of expected and detected deviations

to limit the number of false positive and false negative authentication attempts.

The result of the described holistic design effort is the first practical implementation of biometric authentication in BANs that reflects timing and data uncertainties in the physical and cyber parts of the system.

Cyber physical systems (CPS) integrate a variety of engineering areas such as control, mechanical and computer engineering in a holistic design effort. While interdependencies between the different disciplines are key attributes of CPS design science, little is known about the impact of design decisions of the cyber part on the overall system qualities. To investigate these interdependencies, this paper proposes a simulation-based Design Space Exploration (DSE) framework that considers detailed cyber system parameters such as cache size, bus width, and voltage levels in addition to physical and control parameters of the CPS. We propose an exploration algorithm that surfs the parameter configurations in the cyber physical sub-systems, in order to approximate the Pareto-optimal design points with regards to the trade-os among the design objectives, such as energy consumption and control stability. We apply the proposed framework to a network control system for an inverted-pendulum application. The presented holistic evaluation of the identified Pareto-points reveals the presence of non-trivial trade-os, which are imposed by the control, physical, and detailed cyber parameters. For instance the identified energy and control optimal design points comprise configurations with a wide range of CPU speeds, sample times and cache configuration following non-trivial zig-zag patterns. The proposed framework could identify and manage those trade-os and, as a result, is an imperative rst step to automate the search for superior CSP configurations.

The cyber-physical system (CPS) is a term describing a broad range of complex, multi-disciplinary, physically-aware next generation engineered systems that integrate embedded computing technologies into the physical world. Sensors play an important role in this integration because they provide the data extracted from the physical world for the cyber systems. However, this process is likely to be misled by incorrect data due to sensor faults.

In this dissertation, the main focus is on sensor fault mitigation and achieving high reliability in CPS operations. One of the challenges we tackle is timely event detection in CPS under faulty sensor conditions. In this regard, we examine the falling ball example (FBE) using binary event detectors, a controller, and a camera for timely motion detection and estimation of a falling ball. Another challenge we tackle is satisfying thermal comfort and energy efficiency under faulty sensor conditions in a multi-room building incorporating temperature sensors, controllers, and heating, ventilation, and air conditioning (HVAC) systems. For both cases, we adopt a model-based design (MBD) methodology to analyze the effect of sensor faults on the system outcome. In this regard, we develop well-defined fault and system evaluation models and incorporate them into the traditional CPS model that comprises the cyber, interface (e.g., sensors and actuators) and physical models.

We explore various fault mitigation techniques based on redundancies and temporal-spatial correlations between sensors' data in a holistic design perspective. Furthermore, considering compute demands of CPSs, we introduce the XGRID embedded many-core architecture. XGRID makes use of a novel, FPGA-like, programmable interconnect infrastructure offering scalability and deterministic communication. We further introduce a deployment scenario of XGRID as a use case for thermal control of the multi-room building.

Our findings regarding reliable CPS design show that the physical system attributes can be more dominant than the cyber system attributes on the system outcome. In addition, sensor faults may lead to unsatisfactory system outcome since CPSs heavily rely on sensor readings for decision making. Therefore, the analysis of temporal and spatial correlations between sensor readings helps mitigate sensor faults and enable CPSs to utilize sensors' data more efficiently for decision making.

Cyber-Physical Systems (CPS) are composed of computing devices interacting with physical systems. Model-based design is a powerful methodology in CPS design in the implementation of control systems. For instance, Model Predictive Control (MPC) is typically implemented in CPS applications, e.g., in path tracking of autonomous vehicles. MPC deploys a model to estimate the behavior of the physical system at future time instants for a specific time horizon. Ordinary Differential Equations (ODE) are the most commonly used models to emulate the behavior of continuous-time (non-)linear dynamical systems. A complex physical model may comprise thousands of ODEs that pose scalability, performance and power consumption challenges. One approach to address these model complexity challenges are frameworks that automate the development of model-to-model transformation.

In this dissertation, a state-based model with tunable parameters is proposed to operate as a reconfigurable predictive model of the physical system. Moreover, we propose a run-time switching algorithm that selects the best model using machine learning. We employed a metric that formulates the trade-off between the error and computational savings due to model reduction.

Building statistical models are constrained to having expert knowledge and an actual understanding of the modeled phenomenon or process. Also, statistical models may not produce solutions that are as robust in a real-world context as factors outside the model, like disruptions would not be taken into account. Machine learning models have emerged as a solution to account for the dynamic behavior of the environment and automate intelligence acquisition and refinement. Neural networks are machine learning models, well-known to have the ability to learn linear and nonlinear relations between input and output variables without prior knowledge. However, the ability to efficiently exploit resource-hungry neural networks in embedded resource-bound settings is a major challenge.

Here, we proposed Priority Neuron Network (PNN), a resource-aware neural networks model that can be reconfigured into smaller sub-networks at runtime. This approach enables a trade-off between the model's computation time and accuracy based on available resources. The PNN model is memory efficient since it stores only one set of parameters to account for various sub-network sizes. We propose a training algorithm that applies regularization techniques to constrain the activation value of neurons and assigns a priority to each one. We consider the neuron's ordinal number as our priority criteria in that the priority of the neuron is inversely proportional to its ordinal number in the layer. This imposes a relatively sorted order on the activation values. We conduct experiments to employ our PNN as the predictive model in a CPS application. We can see that not only our technique will resolve the memory overhead of DNN architectures but it also reduces the computation overhead for the training process substantially. The training time is a critical matter especially in embedded systems where many NN models are trained on the fly.

Modern embedded systems are required to work in ever increasing dynamic environments, where predicting the computational load on those systems is intractable. However, timely responses to events have to be provided within precise timing constraints in order to guarantee a required level of performance. Consequently, embedded systems by their very nature exhibit real-time characteristics which impose an additional set of restrictions than those in a typical general purpose system. In addition to the limitations of having to perform to strict timing constraints, most embedded systems are constrained by size, weight, energy consumption and cost limitations. As a result, efficient resource management is a critical aspect in embedded systems that must be considered at multiple architectural levels.

The main objective of this work is to present our work on real-time systems that progress to make the next generation embedded systems more predictable and adaptive to dynamic computational changes. To achieve these goals, this phase of our research has focused on the resource synchronization and adaptive scheduling of real-time embedded applications in uni-processor and multi-core environments. The analysis and experiments show that our resource synchronization protocols outperformed other state-of-the art resource access control protocols used in hierarchical scheduled systems. Implemented in VxWorks and applied to applications used in the aerospace industry response times for hard real-time tasks were improved and deadline misses for hard real-time tasks were substantially reduced.

Conventional control theory has been used in many application domains with great success in the past decades. However, novel solutions are required to cope with the challenges arising from complex interaction of fast growing cyber and physical systems. Specifically, integration of classical control methods with Cyber-Physical System (CPS) design tools is a non-trivial task since those methods have been developed to be used by human experts and are not intended to be part of an automatic design tool. On the other hand, the control problems in emerging Cyber-Physical Systems, such as intelligent transportation and autonomous driving, cannot be addressed by conventional control methods due to the high level of uncertainty, complex dynamic model requirements and operational and safety constraints.

In this dissertation, a holistic CPS design approach is proposed in which the control algorithm is incorporated as a building block in the design tool. The proposed approach facilitates the inclusion of physical variability into the design process and reduces the parameter space to be explored. This has been done by adding constraints imposed by the control algorithm.

Furthermore, Reinforcement Learning (RL) as a replacement for convection control solutions are studied in the emerging domain of intelligent transportation systems. Specifically, dynamic tolling assignments and autonomous intersection management are tackled by the state-of-the-art RL methods, namely, Trust Region Policy Optimization and Finite-Difference Gradient Descent. Additionally, Q-learning is used to improve the performance of an embedded controller using a novel formulation in which cyber-system actions, such as changing control sampling time, is combined with the physical action set of the RL agent. Using the proposed approach, it is shown that the power consumption and computational overhead of the embedded control can be improved.

Finally, to address the current lack of available physical benchmarks, an open physical environment benchmarking framework is introduced. In the proposed framework, various components of a physical environment are captured in a unified repository to enable researchers to define and share standard benchmarks that can be used to evaluate different reinforcement algorithms. They can also share the realized environments via the cloud to enable other groups perform experiments on the actual physical environments instead of currently available simulation-based environments.

Arithmetic expressions are the fundamental building blocks of hardware and software systems. An important problem in computational theory is to decide if two arithmetic expressions are equivalent. However, the general problem of equivalence checking, in digital computers, belongs to the NP Hard class of problems. Moreover, existing general techniques for solving this decision problem are applicable to very simple expressions and impractical when applied to more complex expressions found in programs written in high-level languages. In this paper, we propose a method for solving the arithmetic expression equivalence problem using partial evaluation. In particular, our technique is specifically designed to solve the problem of equivalence checking of arithmetic expressions obtained from high-level language descriptions of hardware/software systems. In our method, we use interval analysis to substantially prune the domain space of arithmetic expressions and limit the evaluation effort to a sufficiently limited set of subspaces. Our results show that the proposed method is fast enough to be of use in practice.