Design and Development of an FPGA-based Distributed Computing Processing Platform

This thesis presents two frameworks- a software framework and a hardware core manager framework- which, together, can be used to develop a processing platform using a distributed system of field-programmable gate array (FPGA) boards. The software framework providesusers with the ability to easily develop applications that exploit the processing power of FPGAs while the hardware core manager framework gives users the ability to configure and interact with multiple FPGA boards and/or hardware cores. This thesis describes the design and development of these frameworks and analyzes the performance of a system that was constructed using the frameworks. The performance analysis included measuring the effect of incorporating additional hardware components into the system and comparing the system to a software-only implementation. This work draws conclusions based on the provided results of the performance analysis and offers suggestions for future work.

StreamCloud: An Elastic Parallel-Distributed Stream Processing Engine

In recent years, applications in domains such as telecommunications, network security or large scale sensor networks showed the limits of the traditional store-then-process paradigm. In this context, Stream Processing Engines emerged as a candidate solution for all these applications demanding for high processing capacity with low processing latency guarantees. With Stream Processing Engines, data streams are not persisted but rather processed on the fly, producing results continuously. Current Stream Processing Engines, either centralized or distributed, do not scale with the input load due to single-node bottlenecks. Moreover, they are based on static configurations that lead to either under or over-provisioning. This Ph.D. thesis discusses StreamCloud, an elastic paralleldistributed stream processing engine that enables for processing of large data stream volumes. Stream- Cloud minimizes the distribution and parallelization overhead introducing novel techniques that split queries into parallel subqueries and allocate them to independent sets of nodes. Moreover, Stream- Cloud elastic and dynamic load balancing protocols enable for effective adjustment of resources depending on the incoming load. Together with the parallelization and elasticity techniques, Stream- Cloud defines a novel fault tolerance protocol that introduces minimal overhead while providing fast recovery. StreamCloud has been fully implemented and evaluated using several real word applications such as fraud detection applications or network analysis applications. The evaluation, conducted using a cluster with more than 300 cores, demonstrates the large scalability, the elasticity and fault tolerance effectiveness of StreamCloud. Resumen En los útimos años, aplicaciones en dominios tales como telecomunicaciones, seguridad de redes y redes de sensores de gran escala se han encontrado con múltiples limitaciones en el paradigma tradicional de bases de datos. En este contexto, los sistemas de procesamiento de flujos de datos han emergido como solución a estas aplicaciones que demandan una alta capacidad de procesamiento con una baja latencia. En los sistemas de procesamiento de flujos de datos, los datos no se persisten y luego se procesan, en su lugar los datos son procesados al vuelo en memoria produciendo resultados de forma continua. Los actuales sistemas de procesamiento de flujos de datos, tanto los centralizados, como los distribuidos, no escalan respecto a la carga de entrada del sistema debido a un cuello de botella producido por la concentración de flujos de datos completos en nodos individuales. Por otra parte, éstos están basados en configuraciones estáticas lo que conducen a un sobre o bajo aprovisionamiento. Esta tesis doctoral presenta StreamCloud, un sistema elástico paralelo-distribuido para el procesamiento de flujos de datos que es capaz de procesar grandes volúmenes de datos. StreamCloud minimiza el coste de distribución y paralelización por medio de una técnica novedosa la cual particiona las queries en subqueries paralelas repartiéndolas en subconjuntos de nodos independientes. Ademas, Stream- Cloud posee protocolos de elasticidad y equilibrado de carga que permiten una optimización de los recursos dependiendo de la carga del sistema. Unidos a los protocolos de paralelización y elasticidad, StreamCloud define un protocolo de tolerancia a fallos que introduce un coste mínimo mientras que proporciona una rápida recuperación. StreamCloud ha sido implementado y evaluado mediante varias aplicaciones del mundo real tales como aplicaciones de detección de fraude o aplicaciones de análisis del tráfico de red. La evaluación ha sido realizada en un cluster con más de 300 núcleos, demostrando la alta escalabilidad y la efectividad tanto de la elasticidad, como de la tolerancia a fallos de StreamCloud.

Distributed collaborative processing in wireless sensor networks with application to target localization and beamforming

Abstract The proliferation of wireless sensor networks and the variety of envisioned applications associated with them has motivated the development of distributed algorithms for collaborative processing over networked systems. One of the applications that has attracted the attention of the researchers is that of target localization where the nodes of the network try to estimate the position of an unknown target that lies within its coverage area. Particularly challenging is the problem of estimating the target’s position when we use received signal strength indicator (RSSI) due to the nonlinear relationship between the measured signal and the true position of the target. Many of the existing approaches suffer either from high computational complexity (e.g., particle filters) or lack of accuracy. Further, many of the proposed solutions are centralized which make their application to a sensor network questionable. Depending on the application at hand and, from a practical perspective it could be convenient to find a balance between localization accuracy and complexity. Into this direction we approach the maximum likelihood location estimation problem by solving a suboptimal (and more tractable) problem. One of the main advantages of the proposed scheme is that it allows for a decentralized implementation using distributed processing tools (e.g., consensus and convex optimization) and therefore, it is very suitable to be implemented in real sensor networks. If further accuracy is needed an additional refinement step could be performed around the found solution. Under the assumption of independent noise among the nodes such local search can be done in a fully distributed way using a distributed version of the Gauss-Newton method based on consensus. Regardless of the underlying application or function of the sensor network it is al¬ways necessary to have a mechanism for data reporting. While some approaches use a special kind of nodes (called sink nodes) for data harvesting and forwarding to the outside world, there are however some scenarios where such an approach is impractical or even impossible to deploy. Further, such sink nodes become a bottleneck in terms of traffic flow and power consumption. To overcome these issues instead of using sink nodes for data reporting one could use collaborative beamforming techniques to forward directly the generated data to a base station or gateway to the outside world. In a dis-tributed environment like a sensor network nodes cooperate in order to form a virtual antenna array that can exploit the benefits of multi-antenna communications. In col-laborative beamforming nodes synchronize their phases in order to add constructively at the receiver. Some of the inconveniences associated with collaborative beamforming techniques is that there is no control over the radiation pattern since it is treated as a random quantity. This may cause interference to other coexisting systems and fast bat-tery depletion at the nodes. Since energy-efficiency is a major design issue we consider the development of a distributed collaborative beamforming scheme that maximizes the network lifetime while meeting some quality of service (QoS) requirement at the re¬ceiver side. Using local information about battery status and channel conditions we find distributed algorithms that converge to the optimal centralized beamformer. While in the first part we consider only battery depletion due to communications beamforming, we extend the model to account for more realistic scenarios by the introduction of an additional random energy consumption. It is shown how the new problem generalizes the original one and under which conditions it is easily solvable. By formulating the problem under the energy-efficiency perspective the network’s lifetime is significantly improved. Resumen La proliferación de las redes inalámbricas de sensores junto con la gran variedad de posi¬bles aplicaciones relacionadas, han motivado el desarrollo de herramientas y algoritmos necesarios para el procesado cooperativo en sistemas distribuidos. Una de las aplicaciones que suscitado mayor interés entre la comunidad científica es la de localization, donde el conjunto de nodos de la red intenta estimar la posición de un blanco localizado dentro de su área de cobertura. El problema de la localization es especialmente desafiante cuando se usan niveles de energía de la seal recibida (RSSI por sus siglas en inglés) como medida para la localization. El principal inconveniente reside en el hecho que el nivel de señal recibida no sigue una relación lineal con la posición del blanco. Muchas de las soluciones actuales al problema de localization usando RSSI se basan en complejos esquemas centralizados como filtros de partículas, mientas que en otras se basan en esquemas mucho más simples pero con menor precisión. Además, en muchos casos las estrategias son centralizadas lo que resulta poco prácticos para su implementación en redes de sensores. Desde un punto de vista práctico y de implementation, es conveniente, para ciertos escenarios y aplicaciones, el desarrollo de alternativas que ofrezcan un compromiso entre complejidad y precisión. En esta línea, en lugar de abordar directamente el problema de la estimación de la posición del blanco bajo el criterio de máxima verosimilitud, proponemos usar una formulación subóptima del problema más manejable analíticamente y que ofrece la ventaja de permitir en¬contrar la solución al problema de localization de una forma totalmente distribuida, convirtiéndola así en una solución atractiva dentro del contexto de redes inalámbricas de sensores. Para ello, se usan herramientas de procesado distribuido como los algorit¬mos de consenso y de optimización convexa en sistemas distribuidos. Para aplicaciones donde se requiera de un mayor grado de precisión se propone una estrategia que con¬siste en la optimización local de la función de verosimilitud entorno a la estimación inicialmente obtenida. Esta optimización se puede realizar de forma descentralizada usando una versión basada en consenso del método de Gauss-Newton siempre y cuando asumamos independencia de los ruidos de medida en los diferentes nodos. Independientemente de la aplicación subyacente de la red de sensores, es necesario tener un mecanismo que permita recopilar los datos provenientes de la red de sensores. Una forma de hacerlo es mediante el uso de uno o varios nodos especiales, llamados nodos “sumidero”, (sink en inglés) que actúen como centros recolectores de información y que estarán equipados con hardware adicional que les permita la interacción con el exterior de la red. La principal desventaja de esta estrategia es que dichos nodos se convierten en cuellos de botella en cuanto a tráfico y capacidad de cálculo. Como alter¬nativa se pueden usar técnicas cooperativas de conformación de haz (beamforming en inglés) de manera que el conjunto de la red puede verse como un único sistema virtual de múltiples antenas y, por tanto, que exploten los beneficios que ofrecen las comu¬nicaciones con múltiples antenas. Para ello, los distintos nodos de la red sincronizan sus transmisiones de manera que se produce una interferencia constructiva en el recep¬tor. No obstante, las actuales técnicas se basan en resultados promedios y asintóticos, cuando el número de nodos es muy grande. Para una configuración específica se pierde el control sobre el diagrama de radiación causando posibles interferencias sobre sis¬temas coexistentes o gastando más potencia de la requerida. La eficiencia energética es una cuestión capital en las redes inalámbricas de sensores ya que los nodos están equipados con baterías. Es por tanto muy importante preservar la batería evitando cambios innecesarios y el consecuente aumento de costes. Bajo estas consideraciones, se propone un esquema de conformación de haz que maximice el tiempo de vida útil de la red, entendiendo como tal el máximo tiempo que la red puede estar operativa garantizando unos requisitos de calidad de servicio (QoS por sus siglas en inglés) que permitan una decodificación fiable de la señal recibida en la estación base. Se proponen además algoritmos distribuidos que convergen a la solución centralizada. Inicialmente se considera que la única causa de consumo energético se debe a las comunicaciones con la estación base. Este modelo de consumo energético es modificado para tener en cuenta otras formas de consumo de energía derivadas de procesos inherentes al funcionamiento de la red como la adquisición y procesado de datos, las comunicaciones locales entre nodos, etc. Dicho consumo adicional de energía se modela como una variable aleatoria en cada nodo. Se cambia por tanto, a un escenario probabilístico que generaliza el caso determinista y se proporcionan condiciones bajo las cuales el problema se puede resolver de forma eficiente. Se demuestra que el tiempo de vida de la red mejora de forma significativa usando el criterio propuesto de eficiencia energética.

Neural Processing Unit based on the Optical Bistable Properties of Semiconductor Laser Amplifiers

This paper analyzes the behavior of a neural processing unit based on the optical bistable properties of semiconductor laser amplifiers. A similar unit to the reported here was previously employed in the simulation of the mammalian retina. The main advantages of the present cell are its larger fan-out and the possibility of different responses according to the light wavelength impinging onto the cell. These properties allow to work with larger structures as well as to obtain different behaviors according to the light characteristics. This new approach gives a possible modeling closer to the real biological configurations. Moreover, a more detailed analysis of the basic cell internal behavior is reported

Context-awareness in task automation services by distributed event processing

Everybody has to coordinate several tasks everyday, usually in a manual manner. Recently, the concept of Task Automation Services has been introduced to automate and personalize the task coordination problem. Several user centered platforms and applications have arisen in the last years, that let their users configure their very own automations based on third party services. In this paper, we propose a new system architecture for Task Automation Services in a heterogeneous mobile, smart devices, and cloud services environment. Our architecture is based on the novel idea to employ distributed Complex Event Processing to implement innovative mixed execution profiles. The major advantage of the approach is its ability to incorporate context-awareness and real-time coordination in Task Automation Services.

Some applications of distributed signal processing

In this work we review some earlier distributed algorithms developed by the authors and collaborators, which are based on two different approaches, namely, distributed moment estimation and distributed stochastic approximations. We show applications of these algorithms on image compression, linear classification and stochastic optimal control. In all cases, the benefit of cooperation is clear: even when the nodes have access to small portions of the data, by exchanging their estimates, they achieve the same performance as that of a centralized architecture, which would gather all the data from all the nodes.

Education Complex "Multi-agent Technologies for Parallel and Distributed Information Processing in Telecommunication Networks"

The paper describes education complex "Multi-agent Technologies for Parallel and Distributed Information Processing in Telecommunication Networks".

Atividade multimodal no córtex sensorial primário de ratos.

The currently accepted model of sensory processing states that different senses are processed in parallel, and that the activity of specific cortical regions define the sensorial modality perceived by the subject. In this work we used chronic multielectrode extracellular recordings to investigate to which extent neurons in the visual and tactile primary cortices (V1 and S1) of anesthetized rats would respond to sensory modalities not traditionaly associated with these cortices. Visual stimulation yielded 87% of responsive neurons in V1, while 82% of S1 neurons responded to tactile stimulation. In the same stimulation sessions, we found 23% of V1 neurons responding to tactile stimuli and 22% of S1 neurons responding to visual stimuli. Our data supports an increasing body of evidence that indicates the existence multimodal processing in primary sensory cortices. Our data challenge the unimodal sensory processing paradigm, and suggest the need of a reinterpretation of the currently accepted model of cortical hierarchy.

Does neural input or processing play a greater role in the magnitude of neuroimaging signals?

An important constraint on how hemodynamic neuroimaging signals such as fMRI can be interpreted in terms of the underlying evoked activity is an understanding of neurovascular coupling mechanisms that actually generate hemodynamic responses. The predominant view at present is that the hemodynamic response is most correlated with synaptic input and subsequent neural processing rather than spiking output. It is still not clear whether input or processing is more important in the generation of hemodynamics responses. In order to investigate this we measured the hemodynamic and neural responses to electrical whisker pad stimuli in rat whisker barrel somatosensory cortex both before and after the local cortical injections of the GABAA agonist muscimol. Muscimol would not be expected to affect the thalamocortical input into the cortex but would inhibit subsequent intra-cortical processing. Pre-muscimol infusion whisker stimuli elicited the expected neural and accompanying hemodynamic responses to that reported previously. Following infusion of muscimol, although the temporal profile of neural responses to each pulse of the stimulus train was similar, the average response was reduced in magnitude by ∼79% compared to that elicited pre-infusion. The whisker-evoked hemodynamic responses were reduced by a commensurate magnitude suggesting that, although the neurovascular coupling relationships were similar for synaptic input as well as for cortical processing, the magnitude of the overall response is dominated by processing rather than from that produced from the thalamocortical input alone.

THE RESONATE-AND-FIRE NEURON: TIME DEPENDENT AND FREQUENCY SELECTIVE NEURONS IN NEURAL NETWORKS

The means through which the nervous system perceives its environment is one of the most fascinating questions in contemporary science. Our endeavors to comprehend the principles of neural science provide an instance of how biological processes may inspire novel methods in mathematical modeling and engineering. The application ofmathematical models towards understanding neural signals and systems represents a vibrant field of research that has spanned over half a century. During this period, multiple approaches to neuronal modeling have been adopted, and each approach is adept at elucidating a specific aspect of nervous system function. Thus while bio-physical models have strived to comprehend the dynamics of actual physical processes occurring within a nerve cell, the phenomenological approach has conceived models that relate the ionic properties of nerve cells to transitions in neural activity. Further-more, the field of neural networks has endeavored to explore how distributed parallel processing systems may become capable of storing memory. Through this project, we strive to explore how some of the insights gained from biophysical neuronal modeling may be incorporated within the field of neural net-works. We specifically study the capabilities of a simple neural model, the Resonate-and-Fire (RAF) neuron, whose derivation is inspired by biophysical neural modeling. While reflecting further biological plausibility, the RAF neuron is also analytically tractable, and thus may be implemented within neural networks. In the following thesis, we provide a brief overview of the different approaches that have been adopted towards comprehending the properties of nerve cells, along with the framework under which our specific neuron model relates to the field of neuronal modeling. Subsequently, we explore some of the time-dependent neurocomputational capabilities of the RAF neuron, and we utilize the model to classify logic gates, and solve the classic XOR problem. Finally we explore how the resonate-and-fire neuron may be implemented within neural networks, and how such a network could be adapted through the temporal backpropagation algorithm.

Neural correlates of face familiarity in institutionally reared children with distinctive, atypical social behavior

Although the impact of early adverse experience on neural processing of face familiarity has been studied, research has not taken into account disordered child behavior. This work compared the neural processing of familiar versus strangers' faces in 47 institutionalized children with a mean age of 54 months to determine the effects of (a) the presence versus absence of atypical social behavior and (b) inhibited versus indiscriminant atypical behavior. Results revealed a pattern of cortical hypoactivation in institutionalized children manifesting atypical social behavior and that inhibited children displayed larger neural response to a caregiver's face than to the stranger's, while indiscriminant children did not discriminate between stimuli. These findings suggest that neural correlates of face familiarity are associated with social functioning in institutionalized children.

The spatio-temporal brain dynamics of processing and integrating sound localization cues in humans.

Interaural intensity and time differences (IID and ITD) are two binaural auditory cues for localizing sounds in space. This study investigated the spatio-temporal brain mechanisms for processing and integrating IID and ITD cues in humans. Auditory-evoked potentials were recorded, while subjects passively listened to noise bursts lateralized with IID, ITD or both cues simultaneously, as well as a more frequent centrally presented noise. In a separate psychophysical experiment, subjects actively discriminated lateralized from centrally presented stimuli. IID and ITD cues elicited different electric field topographies starting at approximately 75 ms post-stimulus onset, indicative of the engagement of distinct cortical networks. By contrast, no performance differences were observed between IID and ITD cues during the psychophysical experiment. Subjects did, however, respond significantly faster and more accurately when both cues were presented simultaneously. This performance facilitation exceeded predictions from probability summation, suggestive of interactions in neural processing of IID and ITD cues. Supra-additive neural response interactions as well as topographic modulations were indeed observed approximately 200 ms post-stimulus for the comparison of responses to the simultaneous presentation of both cues with the mean of those to separate IID and ITD cues. Source estimations revealed differential processing of IID and ITD cues initially within superior temporal cortices and also at later stages within temporo-parietal and inferior frontal cortices. Differences were principally in terms of hemispheric lateralization. The collective psychophysical and electrophysiological results support the hypothesis that IID and ITD cues are processed by distinct, but interacting, cortical networks that can in turn facilitate auditory localization.

Sound objects in time, space and action

SOUND OBJECTS IN TIME, SPACE AND ACTIONThe term "sound object" describes an auditory experience that is associated with an acoustic event produced by a sound source. At cortical level, sound objects are represented by temporo-spatial activity patterns within distributed neural networks. This investigation concerns temporal, spatial and action aspects as assessed in normal subjects using electrical imaging or measurement of motor activity induced by transcranial magnetic stimulation (TMS).Hearing the same sound again has been shown to facilitate behavioral responses (repetition priming) and to modulate neural activity (repetition suppression). In natural settings the same source is often heard again and again, with variations in spectro-temporal and spatial characteristics. I have investigated how such repeats influence response times in a living vs. non-living categorization task and the associated spatio-temporal patterns of brain activity in humans. Dynamic analysis of distributed source estimations revealed differential sound object representations within the auditory cortex as a function of the temporal history of exposure to these objects. Often heard sounds are coded by a modulation in a bilateral network. Recently heard sounds, independently of the number of previous exposures, are coded by a modulation of a left-sided network.With sound objects which carry spatial information, I have investigated how spatial aspects of the repeats influence neural representations. Dynamics analyses of distributed source estimations revealed an ultra rapid discrimination of sound objects which are characterized by spatial cues. This discrimination involved two temporo-spatially distinct cortical representations, one associated with position-independent and the other with position-linked representations within the auditory ventral/what stream.Action-related sounds were shown to increase the excitability of motoneurons within the primary motor cortex, possibly via an input from the mirror neuron system. The role of motor representations remains unclear. I have investigated repetition priming-induced plasticity of the motor representations of action sounds with the measurement of motor activity induced by TMS pulses applied on the hand motor cortex. TMS delivered to the hand area within the primary motor cortex yielded larger magnetic evoked potentials (MEPs) while the subject was listening to sounds associated with manual than non- manual actions. Repetition suppression was observed at motoneuron level, since during a repeated exposure to the same manual action sound the MEPs were smaller. I discuss these results in terms of specialized neural network involved in sound processing, which is characterized by repetition-induced plasticity.Thus, neural networks which underlie sound object representations are characterized by modulations which keep track of the temporal and spatial history of the sound and, in case of action related sounds, also of the way in which the sound is produced.LES OBJETS SONORES AU TRAVERS DU TEMPS, DE L'ESPACE ET DES ACTIONSLe terme "objet sonore" décrit une expérience auditive associée avec un événement acoustique produit par une source sonore. Au niveau cortical, les objets sonores sont représentés par des patterns d'activités dans des réseaux neuronaux distribués. Ce travail traite les aspects temporels, spatiaux et liés aux actions, évalués à l'aide de l'imagerie électrique ou par des mesures de l'activité motrice induite par stimulation magnétique trans-crânienne (SMT) chez des sujets sains. Entendre le même son de façon répétitive facilite la réponse comportementale (amorçage de répétition) et module l'activité neuronale (suppression liée à la répétition). Dans un cadre naturel, la même source est souvent entendue plusieurs fois, avec des variations spectro-temporelles et de ses caractéristiques spatiales. J'ai étudié la façon dont ces répétitions influencent le temps de réponse lors d'une tâche de catégorisation vivant vs. non-vivant, et les patterns d'activité cérébrale qui lui sont associés. Des analyses dynamiques d'estimations de sources ont révélé des représentations différenciées des objets sonores au niveau du cortex auditif en fonction de l'historique d'exposition à ces objets. Les sons souvent entendus sont codés par des modulations d'un réseau bilatéral. Les sons récemment entendus sont codé par des modulations d'un réseau du côté gauche, indépendamment du nombre d'expositions. Avec des objets sonores véhiculant de l'information spatiale, j'ai étudié la façon dont les aspects spatiaux des sons répétés influencent les représentations neuronales. Des analyses dynamiques d'estimations de sources ont révélé une discrimination ultra rapide des objets sonores caractérisés par des indices spatiaux. Cette discrimination implique deux représentations corticales temporellement et spatialement distinctes, l'une associée à des représentations indépendantes de la position et l'autre à des représentations liées à la position. Ces représentations sont localisées dans la voie auditive ventrale du "quoi".Des sons d'actions augmentent l'excitabilité des motoneurones dans le cortex moteur primaire, possiblement par une afférence du system des neurones miroir. Le rôle des représentations motrices des sons d'actions reste peu clair. J'ai étudié la plasticité des représentations motrices induites par l'amorçage de répétition à l'aide de mesures de potentiels moteurs évoqués (PMEs) induits par des pulsations de SMT sur le cortex moteur de la main. La SMT appliquée sur le cortex moteur primaire de la main produit de plus grands PMEs alors que les sujets écoutent des sons associée à des actions manuelles en comparaison avec des sons d'actions non manuelles. Une suppression liée à la répétition a été observée au niveau des motoneurones, étant donné que lors de l'exposition répétée au son de la même action manuelle les PMEs étaient plus petits. Ces résultats sont discuté en termes de réseaux neuronaux spécialisés impliqués dans le traitement des sons et caractérisés par de la plasticité induite par la répétition. Ainsi, les réseaux neuronaux qui sous-tendent les représentations des objets sonores sont caractérisés par des modulations qui gardent une trace de l'histoire temporelle et spatiale du son ainsi que de la manière dont le son a été produit, en cas de sons d'actions.