Página 1 dos resultados de 101 itens digitais encontrados em 0.480 segundos

‣ Distributed computation of fixed points

Fonte: Laboratory for Information and Decision Systems, Massachusetts Institute of Technology Publicador: Laboratory for Information and Decision Systems, Massachusetts Institute of Technology
Formato: 15 leaves; application/pdf
Português
Relevância na Pesquisa
67.85052%
by Dimitri P. Bertsekas.; Bibliography: leaf 15.; "August 1981."; Partial support provided by the Defense Advanced Research Projects Agency under Grant no. ONR-N00014-75-C-1183

‣ Distributed optimization algorithms with communications

Fonte: Laboratory for Information and Decision Systems, Massachusetts Institute of Technology Publicador: Laboratory for Information and Decision Systems, Massachusetts Institute of Technology
Formato: 4 leaves; application/pdf
Português
Relevância na Pesquisa
67.85052%
by John Tsitsiklis and Michael Athans.; Bibliography: leaf 4.; "November, 1983." Caption title.; ONR contract ONR/N00014-77-C-0532 (NR-041-519)

‣ On the complexity of distributed decision problems

Fonte: Laboratory for Information and Decision Systems, Massachusetts Institute of Technology Publicador: Laboratory for Information and Decision Systems, Massachusetts Institute of Technology
Formato: 6 leaves; application/pdf
Português
Relevância na Pesquisa
67.85052%
John Tsitsiklis and Michael Athans.; Bibliography: leaf 6.; "November, 1983." Caption title.; ONR contract ONR/N00014-77-C-0532 (NR-041-519)

‣ Actors : a model of concurrent computation in distributed systems

Fonte: MIT Press Publicador: MIT Press
Formato: ix, 144 p.; 22387931 bytes; application/pdf
Português
Relevância na Pesquisa
68.22871%
Gul A. Agha.; Includes index.; Bibliography: p. 137-140.

‣ Status report of research on distributed information and decision systems in command-and-control; Research on distributed information and decision systems in command-and-control

Fonte: Massachusetts Institute of Technology, Laboratory for Information and Decision Systems Publicador: Massachusetts Institute of Technology, Laboratory for Information and Decision Systems
Formato: ; 1000516 bytes; application/pdf
Português
Relevância na Pesquisa
68.24119%
prepared by: Michael Athans [et al.]; Description based on: Sept.1981/Sept.1982.; Prepared under contract ONR/N00014-77-C-0532 (NR 041-519 and NR 277-300x).

‣ A distributed multiobject tracking algorithm for passive sensor networks

Hughes, Richard P
Fonte: Massachusetts Institute of Technology Publicador: Massachusetts Institute of Technology
Tipo: Tese de Doutorado Formato: 89 leaves; 5103927 bytes; 5103679 bytes; application/pdf; application/pdf
Português
Relevância na Pesquisa
67.976025%
by Richard P. Hughes.; Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 1980.; MICROFICHE COPY AVAILABLE IN ARCHIVES AND ENGINEERING.; Bibliography: leaves 88-89.

‣ 6.852J / 18.437J Distributed Algorithms, Fall 2001; Distributed Algorithms

Lynch, Nancy A. (Nancy Ann), 1948-
Fonte: MIT - Massachusetts Institute of Technology Publicador: MIT - Massachusetts Institute of Technology
Português
Relevância na Pesquisa
78.34079%
Design and analysis of concurrent algorithms, emphasizing those suitable for use in distributed networks. Process synchronization, allocation of computational resources, distributed consensus, distributed graph algorithms, election of a leader in a network, distributed termination, deadlock detection, concurrency control, communication, and clock synchronization. Special consideration given to issues of efficiency and fault tolerance. Formal models and proof methods for distributed computation. Alternate years. From the course home page: Course Description 6.852J / 18.437J intends to: (1) provide a rigorous introduction to the most important research results in the area of distributed algorithms, and (2) prepare interested students to carry out independent research in distributed algorithms. Topics covered include: design and analysis of concurrent algorithms, emphasizing those suitable for use in distributed networks, process synchronization, allocation of computational resources, distributed consensus, distributed graph algorithms, election of a leader in a network, distributed termination, deadlock detection, concurrency control, communication, and clock synchronization. Special consideration is given to issues of efficiency and fault tolerance. Formal models and proof methods for distributed computation are also discussed.

‣ 6.824 Distributed Computer Systems, Fall 2002; Distributed Computer Systems

Morris, Robert Tappan
Fonte: MIT - Massachusetts Institute of Technology Publicador: MIT - Massachusetts Institute of Technology
Português
Relevância na Pesquisa
78.1277%
Abstractions and implementation techniques for design of distributed systems; server design, network programming, naming, storage systems, security, and fault tolerance. Readings from current literature. 6 Engineering Design Points.

‣ Comparação de Algoritmos Paralelos para a Extração de Regras de Associação no Modelo de Memória Distribuída

Mariano, Marcos Alves
Fonte: Universidade Federal de Mato Grosso do Sul Publicador: Universidade Federal de Mato Grosso do Sul
Tipo: Dissertação de Mestrado
Português
Relevância na Pesquisa
68.273076%
Nos ultimos anos, a extra c~ao de conhecimento a partir de grandes volumes de dados t^em sido o objeto de estudo em muitas pesquisas. Com isso, diversas t ecnicas de minera c~ao de dados foram desenvolvidas com o prop osito de descobrir informa c~oes para auxiliar os gestores de empresas e organiza c~oes na tomada de decis~oes. Uma das t ecnicas mais predominantes na minera c~ao de dados e a de extra c~ao de regras de associa c~ao, devido a sua e ci^encia e simplicidade no tratamento das informa c~oes. Com a utiliza c~ao do paralelismo em diversos problemas computacionais, algoritmos paralelos para a minera c~ao de dados foram constru dos utilizando a t ecnica de extra c~ao de regras de associa c~ao. Dentre os algoritmos paralelos mais conhecidos, utilizando o modelo de mem oria distribu da, est a o Apriori, o Eclat e o FP-Growth. Assim, o objetivo deste trabalho e implementar e comparar o desempenho dos algoritmos paralelos Apriori, Eclat e FP-Growth com diferentes n umeros de processadores e tamanhos de bases de dados de entrada.; In the last years, the extraction of knowledge from large amount of data have been the object of study in many surveys. Then, many data mining techniques have been developed in order to discover information to assist managers of companies and organizations in decision-making. One of the most prevalent techniques in data mining is the extraction of association rules...

‣ Semantically annotated multi-protocol adapter nodes: a new approach to implementing network-based information systems using ontologies.

Falkner, Nickolas John Gowland
Fonte: Universidade de Adelaide Publicador: Universidade de Adelaide
Tipo: Tese de Doutorado
Publicado em //2007 Português
Relevância na Pesquisa
68.44124%
Network-based information systems are an important class of distributed systems that serve large and diverse user communities with information and essential network services. Centrally defined standards for interoperation and information exchange ensure that any required functionality is provided but do so at the expense of flexibility and ease of system evolution. This thesis presents a novel approach to implementing network-based information systems in a knowledge-representation-based format using an ontological description of the service. Our approach allows us to provide flexible distributed systems that can conform to global standards while still allowing local developments and protocol extensions. We can share data between systems if we provide an explicit specification of the relationship between the knowledge in the system and the structure and nature of the values shared between systems. Existing distributed systems may share data based on the values and structures of that data but we go beyond syntax-based value exchange to introduce a semantically-based exchange of knowledge. The explicit statement of the semantics and syntax of the system in a machine-interpretable form provides the automated integration of different systems through the use of adapter nodes. Adapter nodes are members of more than one system and seamlessly transport data between the systems. We develop a multi-tier software architecture that characterises the values held inside the system depending on an ontological classification of their structure and context to allow the definition of values in terms of the knowledge that they represent. Initially...

‣ Communication performance measurement and analysis on commodity clusters.

Abdul Hamid, Nor Asilah Wati
Fonte: Universidade de Adelaide Publicador: Universidade de Adelaide
Tipo: Tese de Doutorado
Publicado em //2008 Português
Relevância na Pesquisa
67.98691%
Cluster computers have become the dominant architecture in high-performance computing. Parallel programs on these computers are mostly written using the Message Passing Interface (MPI) standard, so the communication performance of the MPI library for a cluster is very important. This thesis investigates several different aspects of performance analysis for MPI libraries, on both distributed memory clusters and shared memory parallel computers. The performance evaluation was done using MPIBench, a new MPI benchmark program that provides some useful new functionality compared to existing MPI benchmarks. Since there has been only limited previous use of MPIBench, some initial work was done on comparing MPIBench with other MPI benchmarks, and improving its functionality, reliability, portability and ease of use. This work included a detailed comparison of results from the Pallas MPI Benchmark (PMB), SKaMPI, Mpptest, MPBench and MPIBench on both distributed memory and shared memory parallel computers, which has not previously been done. This comparison showed that the results for some MPI routines were significantly different between the different benchmarks, particularly for the shared memory machine. A comparison was done between Myrinet and Ethernet network performance on the same machine...

‣ Scheduling and prototyping of distributed real-time systems (an approach using JINI/JAVASPACES)

Demirtas, Tolga.
Fonte: Monterey, California. Naval Postgraduate School Publicador: Monterey, California. Naval Postgraduate School
Tipo: Tese de Doutorado Formato: xvi, 223 p. : ill. ;
Português
Relevância na Pesquisa
78.25969%
Approved for public release, distribution is unlimited; Scheduling is one of the basic issues in building real-time applications on a distributed computing system. A distributed computing system is typically modeled as a collection of processes interconnected by a communication network. For real-time applications, scheduling is needed to meet applications timing constraints. The major difference between single processor and distributed processors scheduling is that, in addition to deciding which order to execute tasks, distributed processors' scheduling algorithms must also decide which processors the task should run on. Moreover, these algorithms must also take into consideration practical network issues like transmission delay, loss of messages, and synchronization in the absence of a global clock. This thesis proposes a formal model to capture these network constraints and develops a proxy-based network buffer technique to support the inter-process communication for the user-defined distributed real-time systems prototypes generated by the Distributed Computer Aided Prototyping System (DCAPS). The proxy-based technique builds on the Jini/JavaSpaces infrastructure. We have conducted several experiments to measure the response time of inter-process communication via JavaSpaces. We have demonstrated the effectiveness of the proxy-based technique by creating an executable prototype of a user-defined distributed real-time system specification.; Lieutenant Junior Grade...

‣ RITSim: distributed systemC simulation

Cox, David Richard
Fonte: Rochester Instituto de Tecnologia Publicador: Rochester Instituto de Tecnologia
Tipo: Tese de Doutorado Formato: 283470 bytes; 585099 bytes; 111145 bytes; 477 bytes; 2061 bytes; 698 bytes; 5612 bytes; 49 bytes; 283470 bytes; 585099 bytes; application/pdf; application/pdf; text/plain; text/plain; text/plain; application/octet-stream; application/octet-stream; applica
Português
Relevância na Pesquisa
78.1277%
Parallel or distributed simulation is becoming more than a novel way to speedup design evaluation; it is becoming necessary for simulating modern processors in a reasonable timeframe. As architectural features become faster, smaller, and more complex, designers are interested in obtaining detailed and accurate performance and power estimations. Uniprocessor simulators may not be able to meet such demands. The RITSim project uses SystemC to model a processor microarchitecture and memory subsystem in great detail. SystemC is a C++ library built on a discrete-event simulation kernel. Many projects have successfully implemented parallel discrete-event simulation (PDES) frameworks to distribute simulation among several hosts. The field promises significant simulation speedup, possibly leading to faster turnaround time in design space exploration and commercial production. However, parallel implementation of such simulators is not an easy task. It requires modification of the simulation kernel for effective partitioning and synchronization. This thesis explores PDES techniques and presents a distributed version of the SystemC simulation environment. With minimal user interaction, SystemC models can executed on a cluster of workstations using a message-passing library such as the Message Passing Interface (MPI). The implementation is designed for transparency; distribution and synchronization happen with little intervention by the model author. Modification of SystemC is fashioned to promote maintainability with future releases. Furthermore...

‣ Implementation of a module implementor for an activity based distributed system

Mayott, Stewart W.
Fonte: Rochester Instituto de Tecnologia Publicador: Rochester Instituto de Tecnologia
Tipo: Tese de Doutorado
Português
Relevância na Pesquisa
78.168765%
This thesis investigates process environment within a distributed computational network, in particular, an activity-based distributed system. An activity-based distributed system is an object-oriented programming system. Objects are grouped together into logical arrangements called activities as the programmer desires . All program code is contained within objects, and some objects are implementors of object types. Objects in an activity receive activity information related to other objects in the activity and have virtual communication capability with other objects in the same activity. To investigate the feasibility of this environment, a fully functional module implementor was created. Further, other components with limited functionality were coded. An assembly line was simulated within the context of these components. Object death and communication were built into the simulation to investigate this functionality and results upon the system.

‣ A simulation of a distributed file system

Stanley, Alan
Fonte: Rochester Instituto de Tecnologia Publicador: Rochester Instituto de Tecnologia
Tipo: Tese de Doutorado
Português
Relevância na Pesquisa
78.22411%
This thesis presents a simulation of a distributed file system. It is a simplified version of the distributed file system found in the LOCUS distributed operating system. The simulation models a network of multiuser computers of any configuration. The number of sites in the network can range from a minimum of three sites to a maximum of twenty. A simple database management system is supported that allows the creation of an indexed database for reading and updating records. The distributed file system supports a transaction mechanism, record level locking, file replication and update propagation, and network transparency. To test the effect of site failures and network partitioning on the distributed file system, a facility is provided to "crash", "reboot", and "jump to" random sites in the network.

‣ Design and implementation of page based distributed shared memory in distributed database systems

Raman, Padmanabhan
Fonte: Rochester Instituto de Tecnologia Publicador: Rochester Instituto de Tecnologia
Tipo: Tese de Doutorado
Português
Relevância na Pesquisa
78.61173%
This project is the simulation of page based distributed shared memory originally called IVY proposed by Li in 1986[3] and then by Li and Hudak in 1989[4]. The 'Page Based Distributed Shared Memory System' consists of a collection of clients or workstations connected to a server by a Local Area Network. The server contains a shared memory segment within which the distributed database is located. The shared memory segment is divided in the form of pages and hence the name 'Page Based Distributed Shared Memory System' where each page represents a table within that distributed database. In the simplest variant, each page is present on exactly one machine. A reference to a local page is done at full memory speed. An attempt to reference a page on a different machine causes a page fault, which is trapped by the software. The software then sends a message to the remote machine, which finds the needed page and sends it to the requesting process. The fault is then restarted and can now complete, which is achieved with the help of Inter Process Communication (IPC) library. In essence, this design is similar to traditional virtual memory systems: when a process touches a nonresident page, a fault occurs and the operating system fetches the page and maps it in. The difference here is that instead of getting the page from the disk...

‣ Hadoop-cc (collaborative caching) in real time HDFS

Shrivastava, Meenakshi
Fonte: Rochester Instituto de Tecnologia Publicador: Rochester Instituto de Tecnologia
Tipo: Tese de Doutorado
Português
Relevância na Pesquisa
68.607812%
Data is being generated at an enormous rate, due to online activities and use of resources related to computing. To access and handle such enormous amount of data spread, dis- tributed systems is an efficient mechanism. One such widely used distributed filesystem is Hadoop distributed filesystem (HDFS). HDFS follows a cluster approach in order to store huge amounts of data, it is scalable and works on low commodity. It uses MapRe- duce framework to perform analysis and carry computations parallely on these large data sets. Hadoop follows the master/slave architecture decoupling system metadata and appli- cation data where metadata is stored on dedicated server NameNode and application data on DataNodes. In this thesis work, study was performed on Hadoop Architecture, behaviour of filesys- tem and MapReduce in detail and concluded that processing of MapReduce is slow which was further confirmed by initial analysis and experiments performed on default Hadoop configuration. It is known that accessing data from cache is much faster as compared to disk access. Collaborative caching is one such mechanism in which the cache distributed over the clients or dedicated servers or storage devices form a single cache to serve the re- quests. This mechanism helps in improving the performance...

‣ Energy-aware replica selection for data-intensive services in cloud

Li, Bo
Fonte: Rochester Instituto de Tecnologia Publicador: Rochester Instituto de Tecnologia
Tipo: Tese de Doutorado
Português
Relevância na Pesquisa
78.491963%
With the increasing energy cost in data centers, an energy efficient approach to provide data intensive services in the cloud is highly in demand. This thesis solves the energy cost reduction problem of data centers by formulating an energy-aware replica selection problem in order to guide the distribution of workload among data centers. The current popular centralized replica selection approaches address such problems but they lack scalability and are vulnerable to a crash of the central coordinator. Also, they do not take total data center energy cost as the primary optimization target. We propose a simple decentralized replica selection system implemented with two distributed optimization algorithms (consensus-based distributed projected subgradient method and Lagrangian dual decomposition method) to work with clients as a decentralized coordinator. We also compare our energy-aware replica selection approach with the replica selection where a round-robin algorithm is implemented. A prototype of the decentralized replica selection system is designed and developed to collect energy consumption information of data centers. The results show that in the best case scenario of our experiments, the total energy cost using the Lagrangian dual decomposition method is 17.8% less than a baseline round-robin method and 15.3% less than consensus-based distributed projected subgradient method. Also...

‣ Implementation of an activity coordinator for an activity-based distributed system

Shaw, Robert
Fonte: Rochester Instituto de Tecnologia Publicador: Rochester Instituto de Tecnologia
Tipo: Tese de Doutorado
Português
Relevância na Pesquisa
78.243647%
Distributed computing systems offer a number of potential benefits, including: - improved fault-tolerance and reliability - increased processor availability - faster response time - flexibility of system configuration - effective management of geographically distributed resources - integration of special purpose machines into applications In order to realize this potential, support systems that aid in the development of distributed programs are needed. An Activity System facilitates the design and implementation of distributed programs: (1) By allowing the programmer to group functionally related objects into an activity (or job) which is recorded within the system. The information stored concerning relationships between objects may then be used to control their interactions and thus to manage distributed resources. (2) By effectively eliminating the need for the programmer to deal with the underlying details of inter-process communication. The system handles the establishment of communication links between objects in an activity, and controls the routing of messages to activity members. To evaluate the uses of activities in developing distributed programs, I have implemented a portion of such a system; namely, an Activity Coordinator ...

‣ NAMER: A Distributed name server for a connected UNIX Environment

Tatakis, Thomas
Fonte: Rochester Instituto de Tecnologia Publicador: Rochester Instituto de Tecnologia
Tipo: Tese de Doutorado
Português
Relevância na Pesquisa
78.115684%
A naming or identification scheme is crucial to the design of any computer system. One must be able to access and perform operations on resources and data in order to complete a given task. In a distributed system the problem is compounded by the geographical separation of resources. Moreover, the underlying network of computers should be transparent to the users. A user should simply be able to request that an operation be performed and the system should locate the necessary resources. This function of locating a named resource is performed by a special service called a name service. This thesis investigates and develops a name service that provides such capability.