Development of Software Tools for the Test of Ultra Wide Band Receivers

In the last years, the importance of locating people and objects and communicating with them in real time has become a common occurrence in every day life. Nowadays, the state of the art of location systems for indoor environments has not a dominant technology as instead occurs in location systems for outdoor environments, where GPS is the dominant technology. In fact, each location technology for indoor environments presents a set of features that do not allow their use in the overall application scenarios, but due its characteristics, it can well coexist with other similar technologies, without being dominant and more adopted than the others indoor location systems. In this context, the European project SELECT studies the opportunity of collecting all these different features in an innovative system which can be used in a large number of application scenarios. The goal of this project is to realize a wireless system, where a network of fixed readers able to query one or more tags attached to objects to be located. The SELECT consortium is composed of European institutions and companies, including Datalogic S.p.A. and CNIT, which deal with software and firmware development of the baseband receiving section of the readers, whose function is to acquire and process the information received from generic tagged objects. Since the SELECT project has an highly innovative content, one of the key stages of the system design is represented by the debug phase. This work aims to study and develop tools and techniques that allow to perform the debug phase of the firmware of the baseband receiving section of the readers.

Modelling and Verification of a Correction Factor to Evaluate the Efficiency of Solar Thermal Collectors with the Rapid Test Method

A new method for the evaluation of the efficiency of parabolic trough collectors, called Rapid Test Method, is investigated at the Solar Institut Jülich. The basic concept is to carry out measurements under stagnation conditions. This allows a fast and inexpensive process due to the fact that no working fluid is required. With this approach, the temperature reached by the inner wall of the receiver is assumed to be the stagnation temperature and hence the average temperature inside the collector. This leads to a systematic error which can be rectified through the introduction of a correction factor. A model of the collector is simulated with COMSOL Multipyisics to study the size of the correction factor depending on collector geometry and working conditions. The resulting values are compared with experimental data obtained at a test rig at the Solar Institut Jülich. These results do not match with the simulated ones. Consequentially, it was not pos-sible to verify the model. The reliability of both the model with COMSOL Multiphysics and of the measurements are analysed. The influence of the correction factor on the rapid test method is also studied, as well as the possibility of neglecting it by measuring the receiver’s inner wall temperature where it receives the least amount of solar rays. The last two chapters analyse the specific heat capacity as a function of pressure and tem-perature and present some considerations about the uncertainties on the efficiency curve obtained with the Rapid Test Method.

Gyrokinetic modelling of plasma transport and investigation on test particle dynamics

Turbulent plasmas inside tokamaks are modeled and studied using guiding center theory, applied to charged test particles, in a Hamiltonian framework. The equations of motion for the guiding center dynamics, under the conditions of a constant and uniform magnetic field and turbulent electrostatic field are derived by averaging over the fast gyroangle, for the first and second order in the guiding center potential, using invertible changes of coordinates such as Lie transforms. The equations of motion are then made dimensionless, exploiting temporal and spatial periodicities of the model chosen for the electrostatic potential. They are implemented numerically in Python. Fast Fourier Transform and its inverse are used. Improvements to the original Python scripts are made, notably the introduction of a power-law curve fitting to account for anomalous diffusion, the possibility to integrate the equations in two steps to save computational time by removing trapped trajectories, and the implementation of multicolored stroboscopic plots to distinguish between trapped and untrapped guiding centers. The post-processing of the results is made in MATLAB. The values and ranges of the parameters chosen for the simulations are selected based on numerous simulations used as feedback tools. In particular, a recurring value for the threshold to detect trapped trajectories is evidenced. Effects of the Larmor radius, the amplitude of the guiding center potential and the intensity of its second order term are studied by analyzing their diffusive regimes, their stroboscopic plots and the shape of guiding center potentials. The main result is the identification of cases anomalous diffusion depending on the values of the parameters (mostly the Larmor radius). The transitions between diffusive regimes are identified. The presence of highways for the super-diffusive trajectories are unveiled. The influence of the charge on these transitions from diffusive to ballistic behaviors is analyzed.

Ricerca delle vulnerabilità e Penetration Test per Architetture Serverless

Il lavoro di tesi proposto è volto allo studio delle architetture Serverless, strutture che permettono agli sviluppatori di comporre facilmente applicazioni distribuite su molti servizi all’interno di un cloud, senza la necessità di gestire il server. Più nello specifico sono state studiate le FaaS (Function-as-a-Service), un modello di elaborazione cloud basato su eventi in cui il codice viene distribuito in container gestiti dalla piattaforma, e successivamente eseguito on-demand. A seguito di una prima parte di ricerca dello stato dell'arte, l'attenzione si è spostata sulla ricerca di vulnerabilità nel contesto del servizio OpenFaaS, un framework open-source che permette agli sviluppatori di distribuire facilmente funzioni e microservizi. Il deployment è stato fatto faasd, quest’ultimo è una semplificazione di OpenFaaS, usa le stesse componenti e lo stesso ecosistema di OpenFaaS ma usa Containerd al posto di Kubernetes. Dopo una prima fase di installazione e log-in il lavoro si è concentrato nelle varie metodologie di penetration test, nonché di ricerca delle vulnerabilità di sicurezza associate a tale paradigma. In informatica, il penetration test è il processo operativo di analisi o valutazione della sicurezza di un sistema o di una rete, simulando l'attacco di un potenziale utente malintenzionato. Nell'ultima fase sono stati condotti vari tentativi di attacco al sistema OpenFaaS tramite l'ausilio di alcuni tool. Inizialmente è stata fatta un'analisi della rete e del traffico tramite tool quali NMAP e Wireshark, per comprenderne meglio la struttura e come faasd creasse le funzioni a partire dai container. Infine, tramite OpenFaaS sono state create delle funzioni per testare la sicurezza e l'affidabilità di quest'ultima. In particolare, le funzioni indagano all'interno dei container al fine di comprendere la possibilità di eseguire code injection e rilevare possibili dati sensibili nel filesystem dell'immagine Docker così come nelle variabili d'ambiente.