Design and implementation of a testbed for data distribution management

Data Distribution Management (DDM) is a core part of High Level Architecture standard, as its goal is to optimize the resources used by simulation environments to exchange data. It has to filter and match the set of information generated during a simulation, so that each federate, that is a simulation entity, only receives the information it needs. It is important that this is done quickly and to the best in order to get better performances and avoiding the transmission of irrelevant data, otherwise network resources may saturate quickly. The main topic of this thesis is the implementation of a super partes DDM testbed. It evaluates the goodness of DDM approaches, of all kinds. In fact it supports both region and grid based approaches, and it may support other different methods still unknown too. It uses three factors to rank them: execution time, memory and distance from the optimal solution. A prearranged set of instances is already available, but we also allow the creation of instances with user-provided parameters. This is how this thesis is structured. We start introducing what DDM and HLA are and what do they do in details. Then in the first chapter we describe the state of the art, providing an overview of the most well known resolution approaches and the pseudocode of the most interesting ones. The third chapter describes how the testbed we implemented is structured. In the fourth chapter we expose and compare the results we got from the execution of four approaches we have implemented. The result of the work described in this thesis can be downloaded on sourceforge using the following link: https://sourceforge.net/projects/ddmtestbed/. It is licensed under the GNU General Public License version 3.0 (GPLv3).

Isolated electrical microgrids employing renewable energy resources:analysis of the electrification of remote communities in Peru.

This project points out a brief overview of several concepts, as Renewable Energy Resources, Distributed Energy Resources, Distributed Generation, and describes the general architecture of an electrical microgrid, isolated or connected to the Medium Voltage Network. Moreover, the project focuses on a project carried out by GRECDH Department in collaboration with CITCEA Department, both belonging to Universitat Politécnica de Catalunya: it concerns isolated microgrids employing renewable energy resources in two communities in northern Peru. Several solutions found using optimization software regarding different generation systems (wind and photovoltaic) and different energy demand scenarios are commented and analyzed from an electrical point of view. Furthermore, there are some proposals to improve microgrid performances, in particular to increase voltage values for each load connected to the microgrid. The extra costs required by the proposed solutions are calculated and their effect on the total microgrid cost are taken into account; finally there are some considerations about the impact the project has on population and on people's daily life.

Large-scale Network Analysis on Distributed Architectures

Questa dissertazione esamina le sfide e i limiti che gli algoritmi di analisi di grafi incontrano in architetture distribuite costituite da personal computer. In particolare, analizza il comportamento dell'algoritmo del PageRank così come implementato in una popolare libreria C++ di analisi di grafi distribuiti, la Parallel Boost Graph Library (Parallel BGL). I risultati qui presentati mostrano che il modello di programmazione parallela Bulk Synchronous Parallel è inadatto all'implementazione efficiente del PageRank su cluster costituiti da personal computer. L'implementazione analizzata ha infatti evidenziato una scalabilità negativa, il tempo di esecuzione dell'algoritmo aumenta linearmente in funzione del numero di processori. Questi risultati sono stati ottenuti lanciando l'algoritmo del PageRank della Parallel BGL su un cluster di 43 PC dual-core con 2GB di RAM l'uno, usando diversi grafi scelti in modo da facilitare l'identificazione delle variabili che influenzano la scalabilità. Grafi rappresentanti modelli diversi hanno dato risultati differenti, mostrando che c'è una relazione tra il coefficiente di clustering e l'inclinazione della retta che rappresenta il tempo in funzione del numero di processori. Ad esempio, i grafi Erdős–Rényi, aventi un basso coefficiente di clustering, hanno rappresentato il caso peggiore nei test del PageRank, mentre i grafi Small-World, aventi un alto coefficiente di clustering, hanno rappresentato il caso migliore. Anche le dimensioni del grafo hanno mostrato un'influenza sul tempo di esecuzione particolarmente interessante. Infatti, si è mostrato che la relazione tra il numero di nodi e il numero di archi determina il tempo totale.

Computer aided design and finite element modeling to predict bulk mechanical properties of bone scaffolds using triply periodic minimal surfaces (progettazione assistita al calcolatore ed analisi con il metodo degli elementi finiti per predire il comportamento meccanico di scaffolds ossei creati con l'uso di superfici minime periodiche in tre direzioni)

The aim of Tissue Engineering is to develop biological substitutes that will restore lost morphological and functional features of diseased or damaged portions of organs. Recently computer-aided technology has received considerable attention in the area of tissue engineering and the advance of additive manufacture (AM) techniques has significantly improved control over the pore network architecture of tissue engineering scaffolds. To regenerate tissues more efficiently, an ideal scaffold should have appropriate porosity and pore structure. More sophisticated porous configurations with higher architectures of the pore network and scaffolding structures that mimic the intricate architecture and complexity of native organs and tissues are then required. This study adopts a macro-structural shape design approach to the production of open porous materials (Titanium foams), which utilizes spatial periodicity as a simple way to generate the models. From among various pore architectures which have been studied, this work simulated pore structure by triply-periodic minimal surfaces (TPMS) for the construction of tissue engineering scaffolds. TPMS are shown to be a versatile source of biomorphic scaffold design. A set of tissue scaffolds using the TPMS-based unit cell libraries was designed. TPMS-based Titanium foams were meant to be printed three dimensional with the relative predicted geometry, microstructure and consequently mechanical properties. Trough a finite element analysis (FEA) the mechanical properties of the designed scaffolds were determined in compression and analyzed in terms of their porosity and assemblies of unit cells. The purpose of this work was to investigate the mechanical performance of TPMS models trying to understand the best compromise between mechanical and geometrical requirements of the scaffolds. The intention was to predict the structural modulus in open porous materials via structural design of interconnected three-dimensional lattices, hence optimising geometrical properties. With the aid of FEA results, it is expected that the effective mechanical properties for the TPMS-based scaffold units can be used to design optimized scaffolds for tissue engineering applications. Regardless of the influence of fabrication method, it is desirable to calculate scaffold properties so that the effect of these properties on tissue regeneration may be better understood.

Approccio network-based alla discriminant analysis mediante HPC per la ricerca di signature ottimali in dati ad alta dimensionalità

Il presente lavoro di tesi si inserisce nell’ambito della classificazione di dati ad alta dimensionalità, sviluppando un algoritmo basato sul metodo della Discriminant Analysis. Esso classifica i campioni attraverso le variabili prese a coppie formando un network a partire da quelle che hanno una performance sufficientemente elevata. Successivamente, l’algoritmo si avvale di proprietà topologiche dei network (in particolare la ricerca di subnetwork e misure di centralità di singoli nodi) per ottenere varie signature (sottoinsiemi delle variabili iniziali) con performance ottimali di classificazione e caratterizzate da una bassa dimensionalità (dell’ordine di 101, inferiore di almeno un fattore 103 rispetto alle variabili di partenza nei problemi trattati). Per fare ciò, l’algoritmo comprende una parte di definizione del network e un’altra di selezione e riduzione della signature, calcolando ad ogni passaggio la nuova capacità di classificazione operando test di cross-validazione (k-fold o leave- one-out). Considerato l’alto numero di variabili coinvolte nei problemi trattati – dell’ordine di 104 – l’algoritmo è stato necessariamente implementato su High-Performance Computer, con lo sviluppo in parallelo delle parti più onerose del codice C++, nella fattispecie il calcolo vero e proprio del di- scriminante e il sorting finale dei risultati. L’applicazione qui studiata è a dati high-throughput in ambito genetico, riguardanti l’espressione genica a livello cellulare, settore in cui i database frequentemente sono costituiti da un numero elevato di variabili (104 −105) a fronte di un basso numero di campioni (101 −102). In campo medico-clinico, la determinazione di signature a bassa dimensionalità per la discriminazione e classificazione di campioni (e.g. sano/malato, responder/not-responder, ecc.) è un problema di fondamentale importanza, ad esempio per la messa a punto di strategie terapeutiche personalizzate per specifici sottogruppi di pazienti attraverso la realizzazione di kit diagnostici per l’analisi di profili di espressione applicabili su larga scala. L’analisi effettuata in questa tesi su vari tipi di dati reali mostra che il metodo proposto, anche in confronto ad altri metodi esistenti basati o me- no sull’approccio a network, fornisce performance ottime, tenendo conto del fatto che il metodo produce signature con elevate performance di classifica- zione e contemporaneamente mantenendo molto ridotto il numero di variabili utilizzate per questo scopo.

Elastic computing on Cloud resources for the CMS experiment

Nowadays, data handling and data analysis in High Energy Physics requires a vast amount of computational power and storage. In particular, the world-wide LHC Com- puting Grid (LCG), an infrastructure and pool of services developed and deployed by a ample community of physicists and computer scientists, has demonstrated to be a game changer in the efficiency of data analyses during Run-I at the LHC, playing a crucial role in the Higgs boson discovery. Recently, the Cloud computing paradigm is emerging and reaching a considerable adoption level by many different scientific organizations and not only. Cloud allows to access and utilize not-owned large computing resources shared among many scientific communities. Considering the challenging requirements of LHC physics in Run-II and beyond, the LHC computing community is interested in exploring Clouds and see whether they can provide a complementary approach - or even a valid alternative - to the existing technological solutions based on Grid. In the LHC community, several experiments have been adopting Cloud approaches, and in particular the experience of the CMS experiment is of relevance to this thesis. The LHC Run-II has just started, and Cloud-based solutions are already in production for CMS. However, other approaches of Cloud usage are being thought of and are at the prototype level, as the work done in this thesis. This effort is of paramount importance to be able to equip CMS with the capability to elastically and flexibly access and utilize the computing resources needed to face the challenges of Run-III and Run-IV. The main purpose of this thesis is to present forefront Cloud approaches that allow the CMS experiment to extend to on-demand resources dynamically allocated as needed. Moreover, a direct access to Cloud resources is presented as suitable use case to face up with the CMS experiment needs. Chapter 1 presents an overview of High Energy Physics at the LHC and of the CMS experience in Run-I, as well as preparation for Run-II. Chapter 2 describes the current CMS Computing Model, and Chapter 3 provides Cloud approaches pursued and used within the CMS Collaboration. Chapter 4 and Chapter 5 discuss the original and forefront work done in this thesis to develop and test working prototypes of elastic extensions of CMS computing resources on Clouds, and HEP Computing “as a Service”. The impact of such work on a benchmark CMS physics use-cases is also demonstrated.

Deep learning for computer vision: a comparison between convolutional neural networks and hierarchical temporal memories on object recognition tasks

In recent years, Deep Learning techniques have shown to perform well on a large variety of problems both in Computer Vision and Natural Language Processing, reaching and often surpassing the state of the art on many tasks. The rise of deep learning is also revolutionizing the entire field of Machine Learning and Pattern Recognition pushing forward the concepts of automatic feature extraction and unsupervised learning in general. However, despite the strong success both in science and business, deep learning has its own limitations. It is often questioned if such techniques are only some kind of brute-force statistical approaches and if they can only work in the context of High Performance Computing with tons of data. Another important question is whether they are really biologically inspired, as claimed in certain cases, and if they can scale well in terms of "intelligence". The dissertation is focused on trying to answer these key questions in the context of Computer Vision and, in particular, Object Recognition, a task that has been heavily revolutionized by recent advances in the field. Practically speaking, these answers are based on an exhaustive comparison between two, very different, deep learning techniques on the aforementioned task: Convolutional Neural Network (CNN) and Hierarchical Temporal memory (HTM). They stand for two different approaches and points of view within the big hat of deep learning and are the best choices to understand and point out strengths and weaknesses of each of them. CNN is considered one of the most classic and powerful supervised methods used today in machine learning and pattern recognition, especially in object recognition. CNNs are well received and accepted by the scientific community and are already deployed in large corporation like Google and Facebook for solving face recognition and image auto-tagging problems. HTM, on the other hand, is known as a new emerging paradigm and a new meanly-unsupervised method, that is more biologically inspired. It tries to gain more insights from the computational neuroscience community in order to incorporate concepts like time, context and attention during the learning process which are typical of the human brain. In the end, the thesis is supposed to prove that in certain cases, with a lower quantity of data, HTM can outperform CNN.

Proprietà statistiche di network dinamici

L'obbiettivo di questa tesi è quello di studiare alcune proprietà statistiche di un random walk su network. Dopo aver definito il concetto di network e di random walk su network, sono state studiate le caratteristiche dello stato stazionario di questo sistema, la loro dipendenza dalla topologia della rete e l'andamento del sistema verso l'equilibrio, con particolare interesse per la distribuzione delle fluttuazioni delle popolazioni sui differenti nodi, una volta raggiunto lo stato stazionario. In seguito, si è voluto osservare il comportamento del network sottoposto ad una forzatura costante, rappresentata da sorgenti e pozzi applicati in diversi nodi, e quindi la sua suscettività a perturbazioni esterne. Tramite alcune simulazioni al computer, viene provato che una forzatura esterna modifica in modo diverso lo stato del network in base alla topologia di quest'ultimo. Dai risultati si è trovato quali sono i nodi che, una volta perturbati, sono in grado di cambiare ampiamente lo stato generale del sistema e quali lo influenzano in minima parte.

A QoS Controller Framework Compliant with the ETSI Network Function Virtualization Specification

The 5th generation of mobile networking introduces the concept of “Network slicing”, the network will be “sliced” horizontally, each slice will be compliant with different requirements in terms of network parameters such as bandwidth, latency. This technology is built on logical instead of physical resources, relies on virtual network as main concept to retrieve a logical resource. The Network Function Virtualisation provides the concept of logical resources for a virtual network function, enabling the concept virtual network; it relies on the Software Defined Networking as main technology to realize the virtual network as resource, it also define the concept of virtual network infrastructure with all components needed to enable the network slicing requirements. SDN itself uses cloud computing technology to realize the virtual network infrastructure, NFV uses also the virtual computing resources to enable the deployment of virtual network function instead of having custom hardware and software for each network function. The key of network slicing is the differentiation of slice in terms of Quality of Services parameters, which relies on the possibility to enable QoS management in cloud computing environment. The QoS in cloud computing denotes level of performances, reliability and availability offered. QoS is fundamental for cloud users, who expect providers to deliver the advertised quality characteristics, and for cloud providers, who need to find the right tradeoff between QoS levels that has possible to offer and operational costs. While QoS properties has received constant attention before the advent of cloud computing, performance heterogeneity and resource isolation mechanisms of cloud platforms have significantly complicated QoS analysis and deploying, prediction, and assurance. This is prompting several researchers to investigate automated QoS management methods that can leverage the high programmability of hardware and software resources in the cloud.