Despite the fast development of telecommunications systems experienced during the last two decades, much progress is expected in the coming years with the introduction of new solutions capable of delivering fast data-rates and ubiquitous connectivity. However, this development can only happen through the evolution of radio-frequency systems, which should be capable of working at high-power and high-speed. At the same time, the power dissipation of these systems should be minimized. In this dissertation, methods for the characterization and modeling of transistors and power amplifiers are presented, along with the necessary nonlinear measurements techniques. In particular, dynamic electrical effects, originated by the properties of the semiconductor materials and by the system configurations, are investigated. Consequently, these phenomena, which cannot be ignored to obtain the wanted performance, are empirically identified and included in models for Gallium Nitride (GaN) transistors and power amplifiers driven by a dynamic voltage supply.

Choreographies are a relatively new tool for designing distributed systems from a global viewpoint. Moreover, choreographies are also free from deadlocks and race conditions by design. Recent theoretical results defined proper Endpoint Projection (EPP) functions to compile choreographic specifications into their single components. Since EPPs are behavioural preserving, projected systems also enjoy freedom from deadlocks and races by construction. Aim of this PhD is to formalise non-trivial features of distributed systems with choreographies and to translate our theoretical results into the practice of implemented systems. To this purpose, we provide two main contributions. The first contribution tackles one of the most challenging features of distributed development: programming correct and consistent runtime updates of distributed systems. Our solution is a theoretical model of dynamic choreographies that provides a clear definition of which components and behaviours can be updated. We prove that compiled choreographic specifications are correct and consistent after any update. We also refine our theoretical model to provide a finer control over updates. On this refinement, we develop a framework for programming adaptable distributed systems. The second contribution covers one of the main issues of implementing theoretical results on choreographies: formalising the compilation from choreographies to executable programs. There is a sensible departure between the present choreographic frameworks and their theoretical models because their theories abstract communications with synchronisation on names (a la CCS/π-calculus) yet they compile to Jolie programs, an executable language that uses correlation — a renown technology of Service-Oriented Computing — for message routing. Our solution is a theory of Applied Choreographies (AC) that models correlation-based message passing. We pinpoint the key theoretical problems and formalise the principles that developers should follow to obtain correct implementations. Finally, we prove our approach by defining a correct compiler from AC to the calculus behind the Jolie language.

This thesis exhibits the results of my PhD Apprenticeship Program, carried out at the “Marposs S.p.a.” firm, in the electronic research division, and at the Department of Physics and Astronomy of the Bologna University, in the INFN's electronics laboratories of the ATLAS group. During these three years of research, I worked on the development and realization of electronic boards dedicated to flexible data acquisition, designed to be applied in several contexts, that need to share high performance FPGAs and high-speed serial communications. The thesis describes the successful application of high-speed configurable electronic devices to two different fields, firstly developed in the particle physics scenario, and then the industrial measurement of mechanical pieces, reaching the main goal of the PhD Apprenticeship Program. The common denominator is the development of high speed electronics based on FPGAs for demanding data acquisition and data processing applications. The thesis describes the contribution to the luminosity monitor of LHC at CERN and illustrates a multi-camera system developed for automatic inspection of mechanical pieces made by a machine tool. The Apprenticeship Program allowed me to continue my academic course in parallel with my working activity, giving me the opportunity to finalize the project started during my internship and thesis for my master degree. It also allowed me to achieve a higher level in education and training in two different contexts of excellence, i.e. the industrial company and the academic research, where I concretely learned the best technical knowledge. The chance of bringing together two distant worlds was the most enthusiastic aspect of this PhD research. The world of industry and academic research face similar problems but with different points of view and goals. I had the opportunity to explore pure academic research, and also to apply the knowledge acquired in these years to the industrial research.

In a future perspective, the need of mapping an unknown indoor environment, of localizing and retrieving information from objects with zero costs and efforts could be satisfied by the adoption of next 5G technologies. Thanks to the mix of mmW and massive arrays technologies, it will be possible to achieve a higher indoor localization accuracy without relying on a dedicated infrastructure for localization but exploiting that designed for communication purposes. Besides users localization and navigation objectives, mapping and thus, the capability of reconstructing indoor scenarios, will be an important field of research with the possibility of sharing environmental information via crowd-sourcing mechanisms between users. Finally, in the Internet of Things vision, it is expected that people, objects and devices will be interconnected to each other with the possibility of exchanging the acquired and estimated data including those regarding objects identification, positioning and mapping contents. To this end, the merge of RFID, WSN and UWB technologies has demonstrated to be a promising solution. Stimulated by this framework, this work describes different technological and signal processing approaches to ameliorate the localization capabilities and the user awareness about the environment. From one side, it has been focused on the study of the localization and mapping capabilities of multi-antenna systems based on 5G technologies considering different technological issues, as for example those related to the existing available massive arrays. From the other side, UWB-RFID systems relying on passive communication schemes have been investigated in terms of localization coverage and by developing different techniques to improve the accuracy even in presence of NLOS conditions.

Mobile robots gained lots of attention in the last decades. Because of its flexibility and increased capabilities of automation, mobile robots are used in many applications: from domotic, to search and rescue missions, to agriculture, environment protection and many more. The main capability of mobile robots to accomplish a mission is the mobility in the work environment. To move in a certain environment the robots should achieve: guidance, navigation and control. This thesis focuses on guidance and control of mobile robots, with application to certain classes of robots: Vertical Take Off and Landing Unmanned Aerial Vehicles (VTOL UAV) and Differential Wheel robots (DWR). The contribution of this thesis is on modeling and control of the two classes of robots, and on novel strategies of combined control and motion planning for kinodynamic systems. A new approach to model a class of multi-propeller VTOL is proposed, with the aim of generating a general model for a system as a composition of elementary modules such as actuators and payloads. Two control law for VTOL vehicles and DWR are proposed. The goal of the first is to generate a simple yet powerful control to globally asymptotically stabilize a VTOL for acrobatic maneuvers. The second is a simple saturated input control law for trajectory tracking of a DWR model in 2D. About planning, a novel approach to generate non-feasible trajectories for robots that still guarantees a correct path for kinodynamic planning is proposed. The goal is to reduce the runtime of planners to be used in real-time and realistic scenario. Moreover an innovative framework for mobile robots motion planning with the use of Discrete Event Systems theory is introduced. The two proposed approaches allow to build a global, robust, real-time, quasi-optimal, kinodynamic planner suitable for replanning.

This thesis investigates the possibility of integrating the standard CMOS design process with additional microstructures enhancing circuit functionalities. More specifically, the thesis faces the problem of miniaturization of magnetic and piezoelectric devices mostly focused on the application field of EH (Energy Harvesting) systems and ultra-low power and ultra-low voltage systems. It shows all the most critical aspects which have to be taken into account during the design process of miniaturized inductors for PwrSoC (Power System on Chip) or transformers. Furthermore it shows that it is possible to optimize the inductance value and also performances by means of a proper choice of the size of the planar core or choosing a different layout shape such as a serpentine shape in place of the classic toroidal one. A new formula for the correct evaluation of the MPL (Magnetic Path Length) was also introduced. Concerning the piezoelectric counterpart, it is focused on the design and simulation of various MEMS PTs based on a SOI (Silicon on Insulator) structure with AlN (Alluminum Nitride) as active piezoelectric element, in perspective of having a SoC with embedded MEMS devices and circuitry. Furthermore it demonstrates for the first time the use of a PT (Piezoelectric Transformer) for ultra-low voltage EH applications. A new boost oscillator based on a discrete PZT (Lead Zirconate Titanate) PT instead of a MT (Magnetic Transformer) has been modelled and tested on a circuit made up by discrete devices, showing performances comparable to commercial solutions like the LTC3108 from Linear. Furthermore this novel boost oscillator has been designed in a 0.35μm technology by ST Microelectronics, showing better performances as intuitively expected by the developed mathematical model of the entire system.

The continuous technological progress and the constant growing of information flow we observe every day, brought us an urgent need to find a way to defend our data from malicious intruders; cryptography is the field of computer science that deals with security and studies techniques to protect communications from third parties, but in the recent years there has been a crisis in proving the security of cryptographic protocols, due to the exponential increase in the complexity of modeling proofs. In this scenario we study interactions in a typed lambda-calculus properly defined to fit well into the key aspects of a cryptographic proof: interaction, complexity and probability. This calculus, RSLR, is an extension of Hofmann's SLR for probabilistic polynomial time computations and it is perfect to model cryptographic primitives and adversaries. In particular, we characterize notions of context equivalence and context metrics, when defined on linear contexts, by way of traces, making proofs easier. Furthermore we show how to use this techniqe to obtain a proof methodology for computational indistinguishability, a key notion in modern cryptography; finally we give some motivating examples of concrete cryptographic schemes.

This thesis covers two main topics: the design and performance evaluation of Wireless Body Area Networks (WBANs), and the simulation and mathematical modeling of Delay Tolerant Networks (DTNs). Different Medium Access Control (MAC) protocols for WBANs are implemented on dedicated hardware in order to evaluate, through extensive measurement campaigns, the performance of the network in terms of packet loss rate, delay and energy consumption. Novel solutions to cope with body shadowing and to improve the coexistence with other wireless technologies, are presented and evaluated. An analytic model for the CSMA/CA protocol defined in the IEEE 802.15.6 standard is also presented. The benefits of offloading part of the traffic carried by a wireless backbone to a DTN composed of mobile nodes in a urban environment, is also investigated. A more analytic approach, mainly using tools from stochastic geometry and Markov chains theory, is used to develop a mathematical framework for the evaluation of the performance of routing rules for DTNs.

The design and development of wearable inertial sensor systems for health monitoring has garnered a huge attention in the scientific community and the industry during the last years. Such platforms have a typical architecture and common building blocks to enable data collection, data processing and feedback restitution. In this thesis we analyze power optimization techniques that can be applied to such systems. When reducing power consumption in a wearable system, different trade-offs have to be inevitably faced. We thus propose software techniques that span from well known duty cycling, frequency scaling, data compression to new paradigm such as radio triggering, heterogeneous multi-core and context aware power management.

The negotiation between power consumption, performance, programmability, and portability drives all computing industry designs, in particular the mobile and embedded systems domains. Two design paradigms have proven particularly promising in this context: architectural heterogeneity and many-core processors. Parallel programming models are key to effectively harness the computational power of heterogeneous many-core SoC. This thesis presents a set of techniques and HW/SW extensions that enable performance improvements and that simplify programmability for heterogeneous many-core platforms. The thesis contributions cover vertically the entire software stack for many-core platforms, from hardware abstraction layers running on top of bare-metal, to programming models; from hardware extensions for efficient parallelism support to middleware that enables optimized resource management within many-core platforms. First, we present mechanisms to decrease parallelism overheads on parallel programming runtimes for many-core platforms, targeting fine-grain parallelism. Second, we present programming model support that enables the offload of computational kernels within heterogeneous many-core systems. Third, we present a novel approach to dynamically sharing and managing many-core platforms when multiple applications coded with different programming models execute concurrently. All these contributions were validated using STMicroelectronics STHORM, a real embodiment of a state-of-the-art many-core system. Hardware extensions and architectural explorations were explored using VirtualSoC, a SystemC based cycle-accurate simulator of many-core platforms.