High-performance parallel systems for data-intensive computing

Thesis (Ph.D.)--University of Washington, 2016-08

BUILDING EFFICIENT AND COST-EFFECTIVE CLOUD-BASED BIG DATA MANAGEMENT SYSTEMS

In today’s big data world, data is being produced in massive volumes, at great velocity and from a variety of different sources such as mobile devices, sensors, a plethora of small devices hooked to the internet (Internet of Things), social networks, communication networks and many others. Interactive querying and large-scale analytics are being increasingly used to derive value out of this big data. A large portion of this data is being stored and processed in the Cloud due the several advantages provided by the Cloud such as scalability, elasticity, availability, low cost of ownership and the overall economies of scale. There is thus, a growing need for large-scale cloud-based data management systems that can support real-time ingest, storage and processing of large volumes of heterogeneous data. However, in the pay-as-you-go Cloud environment, the cost of analytics can grow linearly with the time and resources required. Reducing the cost of data analytics in the Cloud thus remains a primary challenge. In my dissertation research, I have focused on building efficient and cost-effective cloud-based data management systems for different application domains that are predominant in cloud computing environments. In the first part of my dissertation, I address the problem of reducing the cost of transactional workloads on relational databases to support database-as-a-service in the Cloud. The primary challenges in supporting such workloads include choosing how to partition the data across a large number of machines, minimizing the number of distributed transactions, providing high data availability, and tolerating failures gracefully. I have designed, built and evaluated SWORD, an end-to-end scalable online transaction processing system, that utilizes workload-aware data placement and replication to minimize the number of distributed transactions that incorporates a suite of novel techniques to significantly reduce the overheads incurred both during the initial placement of data, and during query execution at runtime. In the second part of my dissertation, I focus on sampling-based progressive analytics as a means to reduce the cost of data analytics in the relational domain. Sampling has been traditionally used by data scientists to get progressive answers to complex analytical tasks over large volumes of data. Typically, this involves manually extracting samples of increasing data size (progressive samples) for exploratory querying. This provides the data scientists with user control, repeatable semantics, and result provenance. However, such solutions result in tedious workflows that preclude the reuse of work across samples. On the other hand, existing approximate query processing systems report early results, but do not offer the above benefits for complex ad-hoc queries. I propose a new progressive data-parallel computation framework, NOW!, that provides support for progressive analytics over big data. In particular, NOW! enables progressive relational (SQL) query support in the Cloud using unique progress semantics that allow efficient and deterministic query processing over samples providing meaningful early results and provenance to data scientists. NOW! enables the provision of early results using significantly fewer resources thereby enabling a substantial reduction in the cost incurred during such analytics. Finally, I propose NSCALE, a system for efficient and cost-effective complex analytics on large-scale graph-structured data in the Cloud. The system is based on the key observation that a wide range of complex analysis tasks over graph data require processing and reasoning about a large number of multi-hop neighborhoods or subgraphs in the graph; examples include ego network analysis, motif counting in biological networks, finding social circles in social networks, personalized recommendations, link prediction, etc. These tasks are not well served by existing vertex-centric graph processing frameworks whose computation and execution models limit the user program to directly access the state of a single vertex, resulting in high execution overheads. Further, the lack of support for extracting the relevant portions of the graph that are of interest to an analysis task and loading it onto distributed memory leads to poor scalability. NSCALE allows users to write programs at the level of neighborhoods or subgraphs rather than at the level of vertices, and to declaratively specify the subgraphs of interest. It enables the efficient distributed execution of these neighborhood-centric complex analysis tasks over largescale graphs, while minimizing resource consumption and communication cost, thereby substantially reducing the overall cost of graph data analytics in the Cloud. The results of our extensive experimental evaluation of these prototypes with several real-world data sets and applications validate the effectiveness of our techniques which provide orders-of-magnitude reductions in the overheads of distributed data querying and analysis in the Cloud.

Neuromorphic model for sound source segregation

While humans can easily segregate and track a speaker's voice in a loud noisy environment, most modern speech recognition systems still perform poorly in loud background noise. The computational principles behind auditory source segregation in humans is not yet fully understood. In this dissertation, we develop a computational model for source segregation inspired by auditory processing in the brain. To support the key principles behind the computational model, we conduct a series of electro-encephalography experiments using both simple tone-based stimuli and more natural speech stimulus. Most source segregation algorithms utilize some form of prior information about the target speaker or use more than one simultaneous recording of the noisy speech mixtures. Other methods develop models on the noise characteristics. Source segregation of simultaneous speech mixtures with a single microphone recording and no knowledge of the target speaker is still a challenge. Using the principle of temporal coherence, we develop a novel computational model that exploits the difference in the temporal evolution of features that belong to different sources to perform unsupervised monaural source segregation. While using no prior information about the target speaker, this method can gracefully incorporate knowledge about the target speaker to further enhance the segregation.Through a series of EEG experiments we collect neurological evidence to support the principle behind the model. Aside from its unusual structure and computational innovations, the proposed model provides testable hypotheses of the physiological mechanisms of the remarkable perceptual ability of humans to segregate acoustic sources, and of its psychophysical manifestations in navigating complex sensory environments. Results from EEG experiments provide further insights into the assumptions behind the model and provide motivation for future single unit studies that can provide more direct evidence for the principle of temporal coherence.

Avaliação de técnicas de classificação digital de imagens landsat em diferentes padrões de cobertura da terra em Rondônia

Remotely sensed imagery has been widely used for land use/cover classification thanks to the periodic data acquisition and the widespread use of digital image processing systems offering a wide range of classification algorithms. The aim of this work was to evaluate some of the most commonly used supervised and unsupervised classification algorithms under different landscape patterns found in Rondônia, including (1) areas of mid-size farms, (2) fish-bone settlements and (3) a gradient of forest and Cerrado (Brazilian savannah). Comparison with a reference map based on the kappa statistics resulted in good to superior indicators (best results - K-means: k=0.68; k=0.77; k=0.64 and MaxVer: k=0.71; k=0.89; k=0.70 respectively for three areas mentioned). Results show that choosing a specific algorithm requires to take into account both its capacity to discriminate among various spectral signatures under different landscape patterns as well as a cost/benefit analysis considering the different steps performed by the operator performing a land cover/use map. it is suggested that a more systematic assessment of several options of implementation of a specific project is needed prior to beginning a land use/cover mapping job.

On the discovery of business processes orchestration patterns

COORDINSPECTOR is a Software Tool aiming at extracting the coordination layer of a software system. Such a reverse engineering process provides a clear view of the actually invoked services as well as the logic behind such invocations. The analysis process is based on program slicing techniques and the generation of, System Dependence Graphs and Coordination Dependence Graphs. The tool analyzes Common Intermediate Language (CIL), the native language of the Microsoft .Net Framework, thus making suitable for processing systems developed in any .Net Framework compilable language. COORDINSPECTOR generates graphical representations of the coordination layer together with business process orchestrations specified in WSBPEL 2.0

Implementação em hardware reconfigurável de método de separação de dados hiperespetrais

Relatório do Trabalho Final de Mestrado para obtenção do grau de Mestre em Engenharia de Electrónica e Telecomunicações

An efficient scalable RNS architecture for large dynamic ranges

This paper proposes an efficient scalable Residue Number System (RNS) architecture supporting moduli sets with an arbitrary number of channels, allowing to achieve larger dynamic range and a higher level of parallelism. The proposed architecture allows the forward and reverse RNS conversion, by reusing the arithmetic channel units. The arithmetic operations supported at the channel level include addition, subtraction, and multiplication with accumulation capability. For the reverse conversion two algorithms are considered, one based on the Chinese Remainder Theorem and the other one on Mixed-Radix-Conversion, leading to implementations optimized for delay and required circuit area. With the proposed architecture a complete and compact RNS platform is achieved. Experimental results suggest gains of 17 % in the delay in the arithmetic operations, with an area reduction of 23 % regarding the RNS state of the art. When compared with a binary system the proposed architecture allows to perform the same computation 20 times faster alongside with only 10 % of the circuit area resources.

Intrusion tolerant routing with data consensus in wireless sensor networks

Dissertação para obtenção do Grau de Mestre em Engenharia Informática

Processament de música basat en el seu contingut

Projecte de recerca elaborat a partir d’una estada a la National University of Singapore Singapur, entre juliol i octubre del 2007. Donada l'explosió de la música a l'internet i la ràpida expansió de les col•leccions de música digital, un repte clau en l'àrea de la informació musical és el desenvolupament de sistemes de processament musical eficients i confiables. L'objectiu de la investigació proposada ha estat treballar en diferents aspectes de l'extracció, modelatge i processat del contingut musical. En particular, s’ha treballat en l'extracció, l'anàlisi i la manipulació de descriptors d'àudio de baix nivell, el modelatge de processos musicals, l'estudi i desenvolupament de tècniques d'aprenentatge automàtic per a processar àudio, i la identificació i extracció d'atributs musicals d'alt nivell. S’han revisat i millorat alguns components d'anàlisis d'àudio i revisat components per a l'extracció de descriptors inter-nota i intra-nota en enregistraments monofónics d'àudio. S’ha aplicat treball previ en Tempo a la formalització de diferents tasques musicals. Finalment, s’ha investigat el processat d'alt nivell de música basandonos en el seu contingut. Com exemple d'això, s’ha investigat com músics professionals expressen i comuniquen la seva interpretació del contingut musical i emocional de peces musicals, i hem usat aquesta informació per a identificar automàticament intèrprets. S’han estudiat les desviacions en paràmetres com to, temps, amplitud i timbre a nivell inter-nota i intra-nota.

Sistemes encastats en temps real

Resum En l’actualitat, els sistemes electrònics de processament de dades són cada cop més significatius dins del sector industrial. Són moltes les necessitats que sorgeixen en el món dels sistemes d’autentificació, de l’electrònica aeronàutica, d’equips d’emmagatzemament de dades, de telecomunicacions, etc. Aquestes necessitats tecnològiques exigeixen ser controlades per un sistema fiable, robust, totalment dependent amb els esdeveniments externs i que compleixi correctament les restriccions temporals imposades per tal de que realitzi el seu propòsit d’una manera eficient. Aquí és on entren en joc els sistemes encastats en temps real, els quals ofereixen una gran fiabilitat, disponibilitat, una ràpida resposta als esdeveniments externs del sistema, una alta garantia de funcionament i una àmplia possibilitat d’aplicacions. Aquest projecte està pensat per a fer una introducció al món dels sistemes encastats, com també explicar el funcionament del sistema operatiu en temps real FreeRTOS; el qual utilitza com a mètode de programació l’ús de tasques independents entre elles. Donarem una visió de les seves característiques de funcionament, com organitza tasques mitjançant un scheduler i uns exemples per a poder dissenyar-hi aplicacions.

Trinocular System for 3D motion and Dense Stucture Estimation

The relief of the seafloor is an important source of data for many scientists. In this paper we present an optical system to deal with underwater 3D reconstruction. This system is formed by three cameras that take images synchronously in a constant frame rate scheme. We use the images taken by these cameras to compute dense 3D reconstructions. We use Bundle Adjustment to estimate the motion ofthe trinocular rig. Given the path followed by the system, we get a dense map of the observed scene by registering the different dense local reconstructions in a unique and bigger one

Small scale underwater change detection

In this paper, we present a method to deal with the constraints of the underwater medium for finding changes between sequences of underwater images. One of the main problems of underwater medium for automatically detecting changes is the low altitude of the camera when taking pictures. This emphasise the parallax effect between the images as they are not taken exactly at the same position. In order to solve this problem, we are geometrically registering the images together taking into account the relief of the scene

Creating large and accurate mosaics of the mid-atlantic ridge

We present an approach for creating image mosaics using navigation data consisting on 3D position estimates provided by sensors such as LBL available in deep water surveys. A central issue with acoustic 3D positioning is that the accuracy is far too low compositing the images within reasonable accuracy

Identification of geometrically consistent interest points for 3D scene reconstruction

We propose an algorithm that extracts image features that are consistent with the 3D structure of the scene. The features can be robustly tracked over multiple views and serve as vertices of planar patches that suitably represent scene surfaces, while reducing the redundancy in the description of 3D shapes. In other words, the extracted features will off er good tracking properties while providing the basis for 3D reconstruction with minimum model complexity