34 research outputs found
Gestione Avanzata di dati per applicazioni e servizi real-time data intesive
No full textQuesta tesi affronta le problematiche della gestione degli stream di dati per le applicazioni e i servizi real-time data-intensive. Essa presenta i principali obiettivi ottenuti nella attività di ricerca svolta durante il mio Ph.D nell’ambito del Progetto Pegasus(Industria 2015) che ha come scopo principale la costruzione di un Sistema di trasporto Intelligente Avanzato(ITS). Pegasus mira a introdurre una nuova tipologia di ITS che sfrutti i dati provenienti dalle On-Board Units (OBUs), installate sui veicoli, per offrire ai guidatori servizi avanzati volti a migliorare la mobilità urbana quali, per esempio, il monitoraggio e la prevenzione del traffico e degli incidenti stradali, il suggerimento di percorsi alternativi, di parcheggi disponibili, etc.
In generale, dispositivi come smartphone, OBUs e sensori sono capaci di raccogliere informazioni sull’ambiente che li circonda e di trasmetterle. Ciò permette lo sviluppo di nuove applicazioni data-intensive che sfruttino non solo stream di dati in real-time, ma anche dati storici e dati statici. Alcuni esempi sono le applicazioni di monitoraggio (ITS, monitoraggio dell'aria), le applicazioni militari (tracciamento del plotone), le applicazioni di rete (sistemi di rilevazione di intrusioni). Questa tipologia di applicazioni compie interrogazioni sui dati che vanno ben oltre le standard Continuous Query (CQ), supportate dai tradizionali Data Stream Management Systems (DSMSs) perché necessitano di un livello più alto di conoscenza dato proprio dai dati storici e statici. Per rispondere a questi requisiti, vari contributi in letteratura propongono di integrare i DSMS con le funzionalità dei Database Management Systems (DBMSs). Tuttavia, questo approccio non soddisfa completamente i requisiti progettuali e prestazionali richiesti da queste applicazioni.
Questa tesi perciò presenta la prospettiva opposta e pone le fondamenta per rendere un qualsiasi DBMS capace di gestire nativamente stream di dati. A tal fine, introduciamo una nuova tipologia di tabella, la Streaming Table (ST), come una nuova struttura di memorizzazione dove i dati in stream entrano e rimangono memorizzati per un lungo periodo, idealmente per sempre. Le STs presentano un nuovo paradigma d’accesso: scritture continue e letture sia continue che one-time. Per supportare efficacemente sia update che letture a rate elevati, abbiamo inoltre introdotto due nuove tipologie di indici. Una dettagliata serie di risultati sperimentali mostra l’efficacia della tecnologia proposta rispetto allo stato dell’arte attuale.
Nonostante le STs abbiano buone performance, negli scenari con alto workload, come Pegasus, potrebbe essere necessario ridurlo sul sistema centrale e trasferirlo sui dispositivi periferici. Consideriamo il contesto del progetto Pegasus, dove i veicoli sono equipaggiati con dei dispositivi con sensori e sono capaci di processare e trasmettere informazioni al sistema centrale. In questa prospettiva, lontani dal costruire un sistema distribuito, la tesi presenta i risultati sperimentali della nostra ricerca su diverse tecniche di Data Reduction al fine di minimizzare le comunicazioni Vehicle to Infrastructure (V2I). Abbiamo esplorato due categorie di tecniche: le independent techniques, dove i veicoli inviano in maniera completamente autonoma i dati al sistema centrale e le information-need techniques, dove i dati sono inviati al sistema tenendo conto di informazioni aggiuntive che giungono da esso.
Infine, abbiamo integrato queste soluzioni tecnologiche in un middleware comune per gestire dati real-time, storici, statici e per fornire un ampio range di query possibili, supportando sia le CQ che le one-time queries. Il middleware è stato contestualizzato in uno scenario che trae spunto dalle smart city dove vari servizi sono basati sulla disponibilità immediata dei dati. Infine la tesi mostra i risultati promettenti ottenuti da una varietà di prove sperimentali effettuate su di esso.This thesis focuses on Data Management for data intensive real-time applications and services. It presents the main goals achieved during the course of my Ph. D. that was linked to the Pegasus Project, a project funded by Industria 2015 programme and having the overall goal to build an advanced Intelligent Transportation System (ITS). The Pegasus Project aims to introduce a new ITS paradigm that exploits data coming from On-Board Units (OBUs) to provide drivers with advanced services to enhance mobility, i.e traffic congestion prevention and warnings, alternative route prompting, crash monitoring, parking availability, gas station cheapness, aso.
Generally speaking, recent technologies, such as smartphone, OBUs, sensors, aso, are able to collect a large amount of environmental information and to transmit it in different ways. They offer the basis to new modern data-intensive applications that exploit not only live data but also relatively past as well historical information and static data. Some samples of these novel application scenarios are, for instance, monitoring applications (ITS, smart cities), military applications (platoon tracking), network applications (intrusion detection). These kinds of applications show querying needs that go beyond the standard Continuous Queries (CQs) over data stream paradigm supported by traditional Data Stream Management Systems (DSMSs) because they take advantage from an upper level of knowledge given by historical and static data.
In order to cope with these new needs, some authors proposes the idea of moving DSMSs towards the integration of Database Management System (DBMS) functionalities to augment their capabilities. However, this approach does not completely meet the design and performance needs required by this new kind of applications.
This thesis presents the opposite perspective and lay the groundwork for extending DBMSs to natively support streaming facilities. To this end, we introduce a new kind of table, the streaming table, as a persistent structure where streaming data enters and remains stored for a long period, ideally forever. Streaming tables feature a novel access paradigm: continuous writes and one-time as well as continuous reads. Streaming table are equipped with two novel types of indices that efficiently support both update and scan high rates. A detailed experimental evaluation shows the effectiveness of the proposed technology and that no related state-of-the-art approach is able to achieve the same results.
Although streaming tables achieve good performances, in high workload scenarios, like Pegasus, it could be necessary to reduce the workload on centralized systems and move it to neighbour devices. We considered the context of the PEGASUS Project, where vehicles are equipped with sensor-based devices able to compute and communicate to the system information like vehicles’ position and speed. In this perspective, far from building a distributed system, the thesis presents the result of our experimental investigation about the employment of different data reduction techniques to minimize V2I communication in an ITS. We explored two categories of data reduction techniques: independent techniques, where vehicles autonomously send data to the centralized system, and information-need techniques, where data is sent by taking into account additional data received from the centralized system.
Finally we integrate the above technological solutions in a data management middleware that offers the robustness of a common framework to manage real-time/historical/static data and that provides a wide range of query capabilities by efficiently supporting continuous as well as one-time queries. The middleware has been contextualized to scenario coming from the smart city context where various services are based on the prompt availability.
As a proof of its good performances, the thesis shows the very promising results obtained on a variety of traffic conditions
Advanced Data Management for real-time data intensive applications and services
No full textThis work focuses on Data Management for data intensive real-time applications and services in a mobile and pervasive transportation scenario. It presents the main goals achieved in the course of the PEGASUS Project, a project funded by Industria 2015 programme and having the overall goal to build an advanced Intelligent Transportation System (ITS)
Streaming Tables: Native Support to Streaming Data in DBMSs
No full textData stream management systems (DSMSs) are conceived for running continuous queries (CQs) on the most recently streamed data. This model does not completely fit the needs of several modern data-intensive applications that require to manage recent/historical/static data and execute both CQs and OTQs joining such data. In order to cope with these new needs, some DSMSs have moved toward the integration of database management systems (DBMSs) functionalities to augment their capabilities. In this paper we adopt the opposite perspective and we lay the groundwork for extending DBMSs to natively support streaming facilities. To this end, we introduce a new kind of table, the streaming table, as a persistent structure where streaming data enters and remains stored for a long period, ideally forever. Streaming tables feature a novel access paradigm: continuous writes and one-time as well as continuous reads. We present a streaming table implementation and two novel types of indices that efficiently support both update and scan high rates. A detailed experimental evaluation shows the effectiveness of the proposed technology
A Framework for ITS Data Management in a Smart City Scenario
No full textIn this paper we introduce a technological framework to efficiently support data management in a modern Intelligent Transportation System (ITS). The proposed technology enables the efficient storage of a variety of recent/historical/static data and guarantees its effective querying by supporting continuous as well as one-time queries for the delivering of real-time traffic services. The framework also offers a scalable solution for coping with the acquisition of huge volumes of data by employing data reduction techniques in Vehicle-to-Infrastructure transmissions. Experimental evaluation on the Linear Road ITS benchmark and along various simulated scenarios demonstrates that the proposed framework efficiently supports smart city data needs
A Data Management Middleware for ITS Services in Smart Cities
No full textA major societal challenge to be tackled in megacities is sustainable urban transportation. Intelligent Transportation Systems (ITSs) are actually data-centric applications that need to store and query real-time as well as historical/static data from various data sources and have to provide timely responses to users' transportation needs.
In this paper we introduce a data management middleware that offers the robustness of a common framework to support the development of smart applications having the above needs. It supports the efficient storage and access to real-time and historical/static data and provides both one-time and continuous query capabilities. While the middleware has been designed to be general and versatile to support data management for any kind of application, in this paper we explore its suitability to ITS smart services also by means of an experimental evaluation conducted on a variety of traffic scenarios
Evaluation of Data Reduction Techniques for Vehicle to Infrastructure Communication Saving Purposes
No full textIn this paper we investigate the employment of different data reduction techniques to minimize V2I communication in an Intelligent Transportation System (ITS). We consider the context of the PEGASUS Project, where vehicles are equipped with sensor-based devices able to compute and communicate to a Control Centre (CC) information like vehicleśs position and speed. The CC relies on a general-purpose data management module that supports the execution of continuous queries as well as standard SQL one-time queries on the collected data to provide various infomobility services.
The paper explores two categories of data reduction techniques: independent techniques, where vehicles autonomously send data to the CC, and information-need techniques, where data is sent by taking into account additional data received from the CC. The paper discusses and implements the technical changes needed in the CC to support the required info-mobility services under the reduced availability of data. All the investigated techniques have been extensively evaluated in a variety of traffic scenarios
Mitochondrial fission and cristae disruption increase the response of cell models of Huntington's disease to apoptotic stimuli.
No full textHuntington's disease (HD), a genetic neurodegenerative disease caused by a polyglutamine expansion in the Huntingtin (Htt) protein, is accompanied by multiple mitochondrial alterations. Here, we show that mitochondrial fragmentation and cristae alterations characterize cellular models of HD and participate in their increased susceptibility to apoptosis. In HD cells, the increased basal activity of the phosphatase calcineurin dephosphorylates the pro-fission dynamin related protein 1 (Drp1), increasing its mitochondrial translocation and activation, and ultimately leading to fragmentation of the organelle. The fragmented HD mitochondria are characterized by cristae alterations that are aggravated by apoptotic stimulation. A genetic analysis indicates that correction of mitochondrial elongation is not sufficient to rescue the increased cytochrome c release and cell death observed in HD cells. Conversely, the increased apoptosis can be corrected by manoeuvres that prevent fission and cristae remodelling. In conclusion, the cristae remodelling of the fragmented HD mitochondria contributes to their hypersensitivity to apoptosis
Specific structural determinants are responsible for the antioxidant activity and the cell cycle effects of resveratrol
No full textResveratrol (3,4',5-trihydroxy-trans-stilbene) is a natural phytoalexin found in grapes and wine, which shows antioxidant and antiproliferative activities. In this study we have investigated whether these properties are dependent on similar or different structural determinants of the molecule. To this purpose, resveratrol derivatives, in which all or each single hydroxylic function were selectively substituted with methyl groups, were synthesized. Analogues with the stilbenic double bond reduced or with the stereoisomery modified were also investigated. The antioxidant activity of these compounds was evaluated by measuring the inhibition of citronellal thermo-oxidation, or the reduction of 2,2 diphenyl-1-picrylhydrazyl radical. In addition, the protection against lipid peroxidation was determined in rat liver microsomes, and in human primary cell cultures. The antiproliferative activity was evaluated by a clonogenic assay, and by analysis of cell cycle progression and DNA synthesis. The results showed that the hydroxyl group in 4' position is not the sole determinant for antioxidant activity. In contrast, the presence of 4'-OH together with stereoisomery in the trans conformation (4'-hydroxystyryl moiety) was absolutely required for inhibition of cell proliferation. Enzymatic assays in vitro demonstrated that inhibition of DNA synthesis was induced by a direct interaction of resveratrol with DNA polymerases alpha and delta
Inhibition of mammalian DNA polymerases by resveratrol: mechanism and structural determinants.
No full textResveratrol, a natural compound found in many dietary plants and in red wine, plays an important role in the prevention of many human pathological processes, including inflammation, atherosclerosis and carcinogenesis. We have shown that the antiproliferative activity of resveratrol correlated with its ability to inhibit the replicative pols (DNA polymerases) alpha and delta in vitro [Stivala, Savio, Carafoli, Perucca, Bianchi, Maga, Forti, Pagnoni, Albini, Prosperi and Vannini (2001) J. Biol. Chem. 276, 22586-22594]. In this paper, we present the first detailed biochemical investigation on the mechanism of action of resveratrol towards mammalian pols. Our results suggest that specific structural determinants of the resveratrol molecule are responsible for selective inhibition of different mammalian pols, such as the family B pol alpha and the family X pol lambda. Moreover, the resveratrol derivative trans-3,5-dimethoxy-4-hydroxystilbene, which is endowed with a strong antiproliferative activity (Stivala et al., 2001), can inhibit pols alpha and lambda and also suppress the in vitro SV40 DNA replication. The potency of inhibition is similar to that of aphidicolin, an inhibitor of the three replicative pols alpha, delta and epsilon. Our findings establish the necessary background for the synthesis of resveratrol derivatives having more selective and potent antiproliferative activity
