Komplexe Muster neuronaler Aktivität entstehen während der Sharp-wave Ripples (SWRs) im Hippocampus und während der Up States im Neokortex (Zuständen mit hoher Aktivität). Sequenzen von Verhalten, die in der Vergangenheit erlebt wurden, werden während des komplexen Musters abgespielt. Die zugrunde liegenden Mechanismen sind nicht gründlich erforscht: Wie können kleine synaptische Veränderungen die großflächige Netzwerkaktivität während des Gedächtnisabrufes und der Gedächtniskonsolidierung kontrollieren? Im ersten Teil dieser Abhandlung wird die Hypothese aufgestellt, dass eine schwache synaptische Konnektivität zwischen Hebbschen Assemblies von der bereits vorhandenen rekurrenten Konnektivität gefördert wird. Diese Hypothese wird auf folgende Weise geprüft: die vorwärts gekoppelten Assembly-Sequenzen werden in neuronale Netzwerke eingebettet, mit einem Gleichgewicht zwischen exzitatorischer und inhibitorischer Aktivität. Simulationen und analytische Berechnungen haben gezeigt, dass rekurrente Verbindungen innerhalb der Assemblies zu einer schnelleren Signalverstärkung führen, was eine Reduktion der notwendigen Verbindungen zwischen den Assemblies zur Folge hat. Diese Aktivität kann entweder von kleinen sensorisch ähnlichen Inputs hervorgerufen werden oder entsteht spontan infolge von Aktivitätsschwankungen. Globale -- möglicherweise neuromodulatorische -- Änderungen der neuronalen Erregbarkeit können daher die Netzwerkzustände steuern, die Gedächnisabruf und die Konsolidierung begünstigen. Der zweite Teil der Arbeit geht der Herkunft der SWRs nach, die in vitro beobachtet wurden. Neueste Studien haben gezeigt, dass SWR-ähnliche Erscheinungen durch optogenetische Stimulation der Subpopulationen von inhibitorischen Neuronen hervorgerufen werden können (Schlingloff et al., 2014). Um diese Ergebnisse zu erklären wird ein de-inhibierendes Schaltkreis-Modell diskutiert, das die beobachteten Populationsausbrüche generieren kann. Die Auswirkungen der pharmakologischen GABAergischen Modulatoren auf die SWR-Häufigkeit werden in vitro untersucht. Die gewonnenen Ergebnisse wurden in Rahmen des Schaltkreis-Modells analysiert. Insbesondere wird den folgenden Fragen nachgegangen: Wie unterdrückt Gabazine, ein GABA_A-Rezeptor-Antagonist, die Entwicklung von SWRs? Wird das Zeitintervall zwischen SWRs durch die Dynamik der GABA_B Rezeptoren moduliert? / Complex patterns of neural activity appear during up-states in the neocortex and sharp-wave ripples (SWRs) in the hippocampus, including sequences that resemble those during prior behavioral experience. The mechanisms underlying this replay are not well understood. How can small synaptic footprints engraved by experience control large-scale network activity during memory retrieval and consolidation? In the first part of this thesis, I hypothesise that sparse and weak synaptic connectivity between Hebbian assemblies are boosted by pre-existing recurrent connectivity within them. To investigate this idea, sequences of assemblies connected in a feedforward manner are embedded in random neural networks with a balance of excitation and inhibition. Simulations and analytical calculations show that recurrent connections within assemblies allow for a fast amplification of signals that indeed reduces the required number of inter-assembly connections. Replay can be evoked by small sensory-like cues or emerge spontaneously by activity fluctuations. Global--potentially neuromodulatory--alterations of neuronal excitability can switch between network states that favor retrieval and consolidation. The second part of this thesis investigates the origin of the SWRs observed in in-vitro models. Recent studies have demonstrated that SWR-like events can be evoked after optogenetic stimulation of subpopulations of inhibitory neurons (Schlingloff et al., 2014; Kohus et al., 2016). To explain these results, a 3-population model is discussed as a hypothetical disinhibitory circuit that could generate the observed population bursts. The effects of pharmacological GABAergic modulators on the SWR incidence in vitro are analysed. The results are discussed in the light of the proposed disinhibitory circuit. In particular, how does gabazine, a GABA_A receptor antagonist, suppress the generation of SWRs? Another explored question is whether the slow dynamics of GABA_B receptors is modulating the time scale of the inter-event intervals.

Les plates-formes MPSoC permettent de satisfaire les contraintes de performance, de flexibilité et de consommation énergétique requises par les systèmes embarqués émergents. Elles intègrent un nombre important de processeurs, des blocs de mémoire et des périphériques, hiérarchiquement organisés par un réseau d'interconnexion. Le développement du logiciel est réputé difficile, notamment dû à la gestion d'un grand nombre d'entités (tâches/threads/processus). L'exécution concurrente de ces entités permet d'exploiter efficacement l'architecture mais complexifie le processus de mise au point et notamment l'analyse des erreurs. D'une part, les exécutions peuvent être non-déterministes notamment dû à la concurrence, c'est à dire qu'elles peuvent se dérouler d'une manière différente à chaque reprise. En conséquence, il n'est pas garanti qu'une erreur se produirait durant la phase de mise au point. D'autre part, la complexité de l'architecture et de l'exécution peut rendre trop important le nombre d'éléments à analyser afin d'identifier une erreur. Il pourrait donc être difficile de se focaliser sur des éléments potentiellement fautifs. Un des défis majeurs du développement logiciel MPSoC est donc de réduire le temps de la mise au point. Dans cette thèse, nous proposons une méthodologie de mise au point qui aide le développeur à identifier les erreurs dans le logiciel MPSoC. Notre premier objectif est de déboguer une même exécution plusieurs fois afin d'analyser des sources potentielles de l'erreur jusqu'à son identification. Nous avons donc identifié les sources de non-déterminisme MPSoC et proposé des mécanismes de ré-exécution déterministe les plus adaptés. Notre deuxième objectif vise à minimiser les ressources pour reproduire une exécution afin de satisfaire la contrainte MPSoC de maîtrise de l'intrusion. Nous avons donc utilisé des mécanismes efficaces de ré-exécution déterministe et considéré qu'une partie du comportement non-déterministe. Le troisième objectif est de permettre le passage à l'échelle, c'est à dire de déboguer des exécutions caractérisées par un nombre d'éléments de plus en plus croissant. Nous avons donc proposé une méthode qui permet de circonscrire et de déboguer qu'une partie de l'exécution. De plus, cette méthode s'applique aux différents types d'architectures et d'applications MPSoC. / MPSoC platforms provide high performance, low power consumption and flexi-bility required by the emerging embedded systems. They incorporate many proces-sing units, memory blocs and peripherals, hierarchically organized by interconnec-tion network. The software development is known to be difficult, namely due to themanagement of multiple entities (tasks/threads/processes). The concurrent execu-tion of these entities allows to exploit efficiently the architecture but complicatesthe refinement process of the software and especially the debugging activity. Onthe one hand, the executions of the software can be non-deterministic, namely dueto the concurrency, i.e. they perform differently each time. Consequently, thereis no guaranties that an error will occur during the debugging activity. On theother hand, the complexity of the architecture and the execution can increase theelements to be analyzed in the debugging process. As a result, it can be difficultto concentrate on the potentially faulty elements. Therefore, one of the most im-portant challenges in the development process of MPSoC software is to reduce thetime of the refinement process.In this thesis, we propose a new methodology to refine the MPSoC softwarewhich helps the developers to do the debugging activity. Our first objective is tobe able to debug the same execution several times in order to analyze potentialsources of the error. To do so, we identified the sources of non-determinism in theMPSoC software executions and propose the most appropriate methods to recordand replay them. Our second objective is to reduce the execution overhead requi-red by the record mechanisms to limit the intrusiveness which is an importantMPSoC constraint. To accomplish this objective, we consider a part of the non-deterministic behaviour and selected efficient record-replay methods. The thirdobjective is to provide a scalable solution, i.e. to be able to debug more and morecomplex executions, characterized by an increasing number of elements. Therefore,we propose a partial replay method which allows to isolate and debug a fraction ofthe execution elements. Moreover, this method applies to different types of archi-tectures and applications MPSoC.

Submitted by Haroudo Xavier Filho (haroudo.xavierfo@ufpe.br) on 2016-03-11T14:34:26Z No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) DissertacaoIvanFranca.pdf: 1367108 bytes, checksum: 8ceed302b395b606d9ac49b5a05987db (MD5) / Made available in DSpace on 2016-03-11T14:34:26Z (GMT). No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) DissertacaoIvanFranca.pdf: 1367108 bytes, checksum: 8ceed302b395b606d9ac49b5a05987db (MD5) Previous issue date: 2015-08-14 / Os quadros de controle IEEE 802.11 desempenham funções importantes na rede sem fio. Dentre elas estão o controle de acesso ao meio de comunicação, a recuperação de quadros armazenados no Ponto de Acesso e a confirmação do recebimento de blocos de quadros ou de certos tipos de quadros. Apesar da importância dos quadros de controle, eles são vulneráveis a ataques de forjação, manipulação e reinjeção devido a inexistência de mecanismos de proteção. Este trabalho propõe um esquema de segurança para quadros de controle em redes IEEE 802.11 a fim de evitar esses ataques. A proposta se diferencia dos trabalhos relacionados por prover um alto grau de segurança em todos os seus módulos com baixo impacto na vazão da rede. Além disso, a proposta não incorre nas fraquezas que eles possuem na contenção dos ataques de reinjeção e no processo de geração e distribuição de chaves. / IEEE 802.11 control frames play important role in the wireless network. Among them are the medium access control, the retrieving of buffered frames in the Access Point, and the acknowledgment of block of frames or certain types of frames. Despite their importance, control frames remain vulnerable to forging, tampering, and replay attacks due to lack of protection mechanisms. This work proposes a security scheme for IEEE 802.11 control frames to prevent such attacks. Our proposal differs from related work by providing a high level of security in all modules along with low impact on network throughput. Furthermore, the proposal avoid the weaknesses that they have in the restraint the replay attacks and in the key generation and distribution process.

Thesis deals with testing of selected vulnerabilities from the IEC 61850 standard and following design of mitigation measures for selected vulnerabilities. Author simulated vulnerabilities of the GOOSE protocol, NTP attack and attack ona MMS client. Those attacks were GOOSE stNum, GOOSE semantic, GOOSE test bit,GOOSE replay, GOOSE flood, NTP spoofing and MMS password capture. Attacks on protocols GOOSE and MMS were successful, attack on NTP was only partially successful since the device confirmed receiving spoofed time, however it did not change it’s inner clock. Author then designed possible mitigation measures. Tool for automatic testing of selected vulnerabilities, parser for the GOOSE protocol and lightweight multiplatform parser for configuration files were created as well.The outcome of this thesis allows the implementation of lager scale tool for penetration testing of industrial networks as well as it allows implementation of discussed mitigation measures.

This thesis deals with the application of bounded model checking method for self-healing assurance of concurrency related problems. The self-healing is currently interested in the Java programming language. Therefore, it concetrate mainly on the model checker Java PathFinder which is built for handling Java programs. The verification method is implemented like the Record&Replay trace strategy for navigation through a state space and performance bounded model checking from reached state through the use of Record&Replay trace strategy. Java PathFinder was extended by new moduls and interfaces in order to perform the bounded model checking for self-healing assurance. Bounded model checking is applied at the neighbourhood of self-healing.

The testing of software for robot controllers has become increasingly difficult as robotic systems become more complex. As the complexity of the systems increases, the number of hardware systems that the robot relies on also grows. This poses a challenge in testing robot controllers, which is crucial to ensure that robots function safely and effectively in their intended applications. While simulation can be used as a platform for software testing, it is not feasible to simulate everything in a virtual environment, especially when test cases require physical connections to hardware for input and output signals sent to robot controllers. Therefore, the objective of this thesis is to replicate I/O device network communication in order to enhance virtual testing processes. The approach employed involves capturing real-time network traffic, modifying and rebuilding it, and subsequently replaying it. The work examines existing academic research on these approaches and technologies, and investigates the specific challenges in the testing process by conducting research within a company leading globally in industrial robot development. A conceptual model is proposed, and a prototype is developed. The solution demonstrates potential in addressing the current challenges in robot controller testing by enabling network capture, modification, and level 4 network traffic replay. However, experimental results reveal various limitations, such as significant delays in generating responses. Therefore, further research and development are required if the solution is to be implemented in a real-world setting.

Researchers and application designers need repeatable methods to evaluateapplications/systems over WiFi. It is hard to reproduce evaluations overWiFi because of rapidly changing wireless quality over time. In this degreeproject, we present NemFi, a trace-driven emulator for accurately recordingthe WiFi trac and later using it to emulate WiFi links in a repeatable fashion.First, we present the advantages of trace-driven emulation over simulationand experimentation. We capture the uctuating WiFi link conditionsin terms of capacity and losses over time and replay captured behavior forany application running in the emulator. Current record-and-replay techniquesfor web trac and cellular networks do not work for WiFi becauseof their inability to distinguish between WiFi losses and losses due to selfinducedcongestion. They are also lacking other WiFi specic features. Inthe absence of a trace-driven emulator for WiFi, NemFi is also equipped toavoid self-induced packet losses. It is thus capable of isolating WiFi relatedlosses which are then replayed by the NemFi's replay. NemFi's record alsoaddresses the frame aggregation and the eect it has on the actual datatransmission capability over the wireless link. NemFi can record frame aggregation,at all instants of the record phase and later accurately replays theaggregation.Experimental results demonstrate that NemFi is not only accurate inrecording the variable-rate WiFi link but also in capturing cross-trac. NemFialso replays the recorded conditions with considerable accuracy. / Forskare och applikationsdesigners behöver repeterbara metoder för att utvärderaapplikationer och system via WiFi. Det är svårt att reproducera utvärderingar genom WiFi eftersom den trådlösa kvalit´en snabbt förändras över tid. I denna rapport presenterar vi NemFi, en spårstyrd emulator för att noggrant registrera WiFi-trafiken och senare använda den för att emulera WiFi-länkar påett repeterbart sätt. Först presenterar vi fördelarna med spårstyrd emulering jämfört med simulering och experiment. Vi fångar de varierande WiFi förhållanden med avseende påkapacitet och förluster över tid och spelar upp fångat beteende för alla applikationer som körs i emulatorn. Nuvarande inspelning och uppspelningstekniker för webbtrafik och mobilnät fungerar inte för WiFi pågrund av deras oförmåga att skilja mellan WiFi-förluster ochförluster pågrund av självinducerad överbelastning. De saknar ocksåandraWiFi-specifika funktioner. I avsaknad av en spårdriven emulator för WiFi är NemFi ocksåutrustade för att undvika självinducerade paketförluster. Den kan alltsåisolera WiFi-relaterade förluster som sedan spelas upp igen av NemFi: s uppspelning. NemFi adresserar ocksåramaggregering och det är effekten påfaktiska dataöverföringsförmåga via den trådlösa länken. NemFi kan spela in ramsamling, vid alla ögonblick i inspelningsfasen och ersätter senare noggrant aggregeringen.Experimentella resultat visar att NemFi inte bara är användbart när det gäller att registrera WiFi-länken med variabel hastighet, utan ocksåför att fånga tvärgående trafik. NemFi ersätter ocksåde inspelade förhållandena medbetydande noggrannhet.

The Internet of things (IoT) device has been widely used in various fields, and its market is expanding rapidly. However, the growing usage of IoT devices also brings more security concerns. The smart door lock is one of the smart home IoT devices that need to be designed securely. This thesis work aims to evaluate and investigate the security aspect of the newest smart door lock. This thesis first provides an introduction and background of penetration testing and creates the threat model. Based on the threat model, some testings are conducted, including state consistency, Man-In-The-Middle (MITM) attack, replay attack, reverse engineering, GPS spoofing, Denial of service (DoS) attack. The result indicates that penetration tests reveal some security problems on the tested device, especially in the access log, traffic between application and server, and the ability of resistance disruption on the WiFi access point. / IoT-enheten har använts i stor utsträckning inom olika områden och dess marknad expanderar snabbt. Den ökande användningen av IoT-enheter medför dock också fler säkerhetsproblem. Det smarta dörrlåset är en av de smarta hem IoT-enheterna som måste utformas säkert. Detta examensarbete syftar till att utvärdera och undersöka säkerhetsaspekten av det nyaste smarta dörrlåset. Denna avhandling ger först en introduktion och bakgrund av penetrationstestning och skapar hotmodellen. Baserat på hotmodellen genomförs vissa tester, inklusive tillståndskonsistens, MITM attack, replay attack, reverse engineering, GPS spoofing, DoS attack. Resultatet indikerar att penetrationstester avslöjar vissa sårbarheter på den testade enheten, särskilt i åtkomstloggen, trafik mellan applikation och server och förmågan till motståndsavbrott på WiFi-åtkomstpunkten.

Graphs are often considered an excellent way of modeling complex real-world problems since they allow to capture relationships between items. Because of their ubiquity, graph embedding techniques have occupied research groups, seeking how vertices can be encoded into a low-dimensional latent space, useful to then perform machine learning. Recently Graph Neural Networks (GNN) have dominated the space of embeddings generation due to their inherent ability to encode latent node dependencies. Moreover, the newly introduced Inductive Graph Neural Networks gained much popularity for inductively learning and representing node embeddings through neighborhood aggregate measures. Even when an entirely new node, unseen during training, appears in the graph, it can still be properly represented by its neighboring nodes. Although this approach appears suitable for dynamic graphs, available systems and training methodologies are agnostic of dynamicity and solely rely on re-processing full graph snapshots in batches, an approach that has been criticized for its high computational costs. This work provides a thorough solution to this particular problem via an efficient prioritybased method for selecting rehearsed samples that guarantees low complexity and high accuracy. Finally, a data-parallel inference method has been evaluated at scale using Apache Flink, a data stream processor for real-time predictions on high volume graph data streams. / Molti problemi nel mondo reale possono essere rappresentati come grafi poichè queste strutture dati consentono di modellare relazioni tra elementi. A causa del loro vasto uso, molti gruppi di ricerca hanno tentato di rappresentare i vertici in uno spazio a bassa dimensione, utile per poi poter utilizzare tecniche di apprendimento automatico. Le reti neurali per grafi sono state ampiamente utilizzate per via della loro capacità di codificare dipendenze tra vertici. Le reti neurali induttive recentemente introdotte, inoltre, hanno guadagnato popolarità poichè consentono di generare rappresentazioni di vertici aggregando altri vertici. In questo modo anche un nodo completamente nuovo può comunque essere rappresentato utilizzando i suoi nodi vicini. Sebbene questo approccio sia adatto per grafici dinamici, i sistemi ad oggi disponibili e gli algoritmi di addestramento si basano esclusivamente sulla continua elaborazione di grafi statici, un approccio che è stato criticato per i suoi elevati costi di calcolo. Questa tesi fornisce una soluzione a questo problema tramite un metodo efficiente per l’allenamento di reti neurali induttive basato su un’euristica per la selezione dei vertici. Viene inoltre descritto un metodo per eseguire predizioni in modo scalabile in tempo reale utilizzando Apache Flink, un sistema per l’elaborazione di grandi quantità di flussi di dati in tempo reale. / Grafer anses ofta vara ett utmärkt sätt att modellera komplexa problem i verkligheten eftersom de gör det möjligt att fånga relationer mellan objekt. På grund av deras allestädes närhet har grafinbäddningstekniker sysselsatt forskningsgrupper som undersöker hur hörn kan kodas in i ett lågdimensionellt latent utrymme, vilket är användbart för att sedan utföra maskininlärning. Nyligen har Graph Neural Networks (GNN) dominerat utrymmet för inbäddningsproduktion tack vare deras inneboende förmåga att koda latenta nodberoenden. Dessutom fick de nyinförda induktiva grafiska nervnäten stor popularitet för induktivt lärande och representerande nodbäddningar genom sammanlagda åtgärder i grannskapet. Även när en helt ny nod, osynlig under träning, visas i diagrammet, kan den fortfarande representeras ordentligt av dess angränsande noder. Även om detta tillvägagångssätt tycks vara lämpligt för dynamiska grafer, är tillgängliga system och träningsmetodologier agnostiska för dynamik och förlitar sig bara på att behandla fullständiga ögonblicksbilder i partier, en metod som har kritiserats för dess höga beräkningskostnader. Detta arbete ger en grundlig lösning på detta specifika problem via en effektiv prioriteringsbaserad metod för att välja repeterade prover som garanterar låg komplexitet och hög noggrannhet. Slutligen har en dataparallell inferensmetod utvärderats i skala med Apache Flink, en dataströmprocessor för realtidsprognoser för grafiska dataströmmar med hög volym.

Emerging machine learning (ML) techniques are showing great potential in prediction performance. However, research and development is often conducted in an environment with extensive computational resources and blinded by prediction performance. In reality, computational resources might be contained on constrained hardware where energy and memory consumption must be restrained. Furthermore, shortages of sufficiently large datasets for ML is a frequent problem, combined with the cost of data retention. This generates a significant demand for sustainable ML. With sustainable ML, practitioners can train ML models on less data, which reduces memory and energy consumption during the training process. To explore solutions to these problems, this thesis dives into several techniques that have been introduced in the literature to achieve energy-savings when training machine learning models. These techniques include Quantization-Aware Training, Model Distillation, Quantized Distillation, Continual Learning and a deeper dive into Siamese Neural Networks (SNNs), one of the most promising techniques for sustainability. Empirical evaluations are conducted using several datasets to illustrate the potential of these techniques and their contribution to sustainable ML. The findings of this thesis show that the energy-saving techniques could be leveraged in some cases to make machine learning models more manageable and sustainable whilst not compromising significant model prediction performance. In addition, the deeper dive into SNNs shows that SNNs can outperform standard classification networks, under both the standard multi-class classification case and the Continual Learning case, whilst being trained on significantly less data. / Maskininlärning har i den senaste tidens forskning visat stor potential och hög precision inom klassificering. Forskning, som ofta bedrivs i en miljö med omfattande beräkningsresurser, kan lätt bli förblindad av precision. I verkligheten är ofta beräkningsresurser lokaliserade på hårdvara där energi- och minneskapacitet är begränsad. Ytterligare ett vanligt problem är att uppnå en tillräckligt stor datamängd för att uppnå önskvärd precision vid träning av maskininlärningsmodeller. Dessa problem skapar en betydande efterfrågan av hållbar maskininlärning. Hållbar maskininlärning har kapaciteten att träna modeller på en mindre datamängd, vilket minskar minne- och energiförbrukning under träningsprocessen. För att utforska hållbar maskininlärning analyserar denna avhandling Quantization-Aware Training, Model Distillation, Quantized Distillation, Continual Learning och en djupare evaluering av Siamesiska Neurala Nätverk (SNN), en av de mest lovande teknikerna inom hållbar maskininlärning. Empiriska utvärderingar utfördes med hjälp av flera olika datamängder för att illustrera potentialen hos dessa tekniker. Resultaten visar att energibesparingsteknikerna kan utnyttjas för att göra maskininlärningsmodeller mer hållbara utan att kompromissa för precision. Dessutom visar undersökningen av SNNs att de kan överträffa vanliga neurala nätverk, med och utan Continual Learning, även om de tränas på betydligt mindre data.