Synchronization of data stream processing has a significant impact on performance of systems where processing of long sequences of data items needs to be done simultaneously. In earlier works on stream processing, syn...
详细信息
ISBN:
(纸本)9780889866379
Synchronization of data stream processing has a significant impact on performance of systems where processing of long sequences of data items needs to be done simultaneously. In earlier works on stream processing, synchronization has been discussed to a limited extent or has been completely overlooked. This work describes a formal model of synchronization in a data stream processing network. We use a notation of data stream processing networks to identify circumstances that necessitate synchronization and express processing of data items in a data stream processing network in terms of database transactions. A technique similar to timestamp ordering of database transactions is used to solve the problems. A solution is presented as a set of rules that govern processing of individual data items. A proof of correctness for the strategy used is also provided.
Recent expectations regarding a new generation of the Web strongly depend on a success of Semantic Web technology. Resource Description Framework (RDF)(1) is the basis for explicit and machine-readable representations...
详细信息
ISBN:
(纸本)9780889866379
Recent expectations regarding a new generation of the Web strongly depend on a success of Semantic Web technology. Resource Description Framework (RDF)(1) is the basis for explicit and machine-readable representations of semantics of various Web resources and enables a framework for interoperability of future Semantic Web-based applications. However it has been pointed out that RDF is not suitable for describing highly dynamic and proactive resources (e.g. industrial devices, processes, etc.). Therefore, an appropriate extension of the existing RDF is necessary. This paper presents the Proactivity Layer of the Smart Resource in Semantic Web with the Resource Agent Behaviour definition. Process performance strategies and coordination methods of such proactive goal-driven resources are considered.
When several independent organizations, whose missions partially overlap, agree to gather together to collaborate in order to achieve a common goal, the resulting collaboration is often referred to in the literature a...
详细信息
ISBN:
(纸本)088986568X
When several independent organizations, whose missions partially overlap, agree to gather together to collaborate in order to achieve a common goal, the resulting collaboration is often referred to in the literature as virtual organization (VO). In this paper we propose a flexible and non-invasive access control infrastructure for VOs that enables member organizations to gain the advantages of the collaboration while preserving their authonomy. The infrastructure is described from the users' point of view and with reference to the European grid project EGEE [1].
An ant colony optimization method for searching in (possibly dynamic and/or unstructured) distributed datasets, as introduced by Jovanovic et. al [1], is considered. This paper provides two new results. Firstly, it de...
详细信息
ISBN:
(纸本)9780889866379
An ant colony optimization method for searching in (possibly dynamic and/or unstructured) distributed datasets, as introduced by Jovanovic et. al [1], is considered. This paper provides two new results. Firstly, it describes how this method can easily be controlled by using different kinds of ants for aggregation of data found: "classic" pheromone aggregation ants should be used if network load caused by a distributed search should be strictly kept within given limits, while one-time aggregation ants should be used if the search process should react quickly due to changes in a dynamic distributed dataset. Secondly, it demonstrates that one-time aggregation ants are more effective than pheromone aggregation ants.
This paper presents a novel approach to reduce the communication overhead incurred when performing multiple multicasts on wormhole-routed three-dimensional mesh and torus networks. Multicasting is a common collective ...
详细信息
ISBN:
(纸本)9780889866386
This paper presents a novel approach to reduce the communication overhead incurred when performing multiple multicasts on wormhole-routed three-dimensional mesh and torus networks. Multicasting is a common collective communication operation implemented on most existing systems via a series of unicast messages. For a single multicast, the transmission time is dominated by the message start-up time. However, in the presence of multiple multicasts the communication and blocking delays become significant. This approach uses a hybrid static-dynamic technique to reduce the communication costs for both single and multiple multicasts. For a single multicast, this technique performs as well as conventional methods. For multiple multicasts, results show that it provides considerable improvement when compared with baseline techniques.
In a distributed system, process synchronization is an important agenda. One of the major duties for process synchronization is mutual exclusion. This paper presents a new centralized fault tolerant distributed mutual...
详细信息
ISBN:
(纸本)9780889866379
In a distributed system, process synchronization is an important agenda. One of the major duties for process synchronization is mutual exclusion. This paper presents a new centralized fault tolerant distributed mutual exclusion algorithm based on Agrawala and El-Abbadi's algorithm. In new algorithm, once coordinator crashes, algorithm can recover lost data and return the coordinator in earlier situation. Thus fault tolerance will ascend and centralize algorithm's "single point of failure" will be omitted. So based algorithm will be more reliable. The only trade off is consuming some inappreciable time in case of coordinator's crash.
In this paper we make use of the LoPC model, which is inspired by both the LogP/PG and BSP models but accounts for contention for message processing resources in parallel programming models to derive a general estimat...
详细信息
ISBN:
(纸本)9780889866379
In this paper we make use of the LoPC model, which is inspired by both the LogP/PG and BSP models but accounts for contention for message processing resources in parallel programming models to derive a general estimate of execution cost. We carry out this cost estimate analysis for three dominate programming models: message passing, shared memory and distributed shared memory. We analyze a typical application;GUPs written in these programming models that have irregular receiver-initiated synchronous communication. The LoPC estimate for this application is shown to be accurate when compared against measured values of runtime of actual empirical computations on an SGI O2000 multiprocessor machine.
Current synchronization engines are mainly designed to reconcile data repositories between multiple clients and a central server on a star-like topology. A different approach is needed to achieve synchronization on pe...
详细信息
ISBN:
(纸本)9780889866379
Current synchronization engines are mainly designed to reconcile data repositories between multiple clients and a central server on a star-like topology. A different approach is needed to achieve synchronization on peer-to-peer topologies where any node can be both client and server and updates may happen independently. Version vectors are one solution to the problem, ensuring global convergence of the datasets and providing straightforward conflict detection, while letting applications to control the conflict resolution semantics in their specific domain. In this paper an implementation of a synchronization engine for contact data in mobile devices using version vectors is presented. The engine is capable of optimistically synchronizing databases among many nodes in a peer-to-peer fashion.
The study of scale-free networks typically assumes that either nodes are homogenously distributed or one strong node monopolizes the entire network. While this makes the analysis trackable, it is not appropriate for s...
详细信息
ISBN:
(纸本)9780889866386
The study of scale-free networks typically assumes that either nodes are homogenously distributed or one strong node monopolizes the entire network. While this makes the analysis trackable, it is not appropriate for scenarios, such as terrorist networks, in which nodes play significant different roles in the network. In this paper, we classify nodes into four types by the roles they play. Nodes from different types are assigned various fitness values. The presence of varied fitness values influences the probability of a node to receive preferentially links from other nodes. Thus our model involves simultaneous interaction between nodes of varied strengths. Although the model allocates the same fitness strength to all nodes of the same type, analytical results suggest that it is sufficient to demonstrate the two key features of scale-free networks: power-law distribution of degree and preferential attachment of new nodes. Simulations are performed to verify the proposed model.
KDDML is a middleware language and system for knowledge discovery. The challenge that motivated the development of a distributed version of the originally "standalone" KDDML (KDD Markup Language) environment...
详细信息
ISBN:
(纸本)088986568X
KDDML is a middleware language and system for knowledge discovery. The challenge that motivated the development of a distributed version of the originally "standalone" KDDML (KDD Markup Language) environment was on one side to exploit the parallelism, and on the other side to overcome the problem of data immovability, a quite frequent restriction on real-world data collections that has principally a privacy-preserving purpose. The last question is addressed by moving the code and "mining" the data "on the place", that is by adapting the computation to the availability and localization of the data.
暂无评论