The DARPA Communicator program has fuelled the design and development of impressive human language technology applications. Its distributed framework has offered numerous benefits to the research community, including ...
详细信息
ISBN:
(纸本)9780780397538
The DARPA Communicator program has fuelled the design and development of impressive human language technology applications. Its distributed framework has offered numerous benefits to the research community, including reduced prototype development time, sharing of components across sites, and provision of a standard evaluation platform. It has also enabled development of client-server applications with complex inter-process communication between modules. However, this latter feature, though beneficial, introduces complexities which reduce overall system robustness to failure. In addition, the ability to handle multiple users and multiple applications from a common interface is not innately supported In this paper, we describe our enhancements to the original Communicator architecture to address robustness issues and to support a multiple multi-user application capability. These enhancements have been evaluated using a series of experiments and they have shown a 7.2% improvement in the robustness of the system. These enhancements are available in our public domain toolkit.
A new secret-sharing-based e-auction scheme is proposed. distributed bid opening is employed to protect bid privacy. It can achieve all the desired properties for sealed-bid auctions at a reasonable cost. Moreover, at...
详细信息
This paper presents an operational semantics of the repetitive model of computation, which is the basis for the repetitive structure modeling (RSM) package defined in the standard UML Marte profile. It also deals with...
详细信息
A grid refinement scheme has been developed for tetrahedral and triangular grid-based calculations in message-passing environments. The element adaption scheme is based on edge bisection of elements marked for refinem...
详细信息
ISBN:
(纸本)354067442X
A grid refinement scheme has been developed for tetrahedral and triangular grid-based calculations in message-passing environments. The element adaption scheme is based on edge bisection of elements marked for refinement by an appropriate error indicator. Hash table/linked list data structures are used to store nodal and element information. The grid along inter-processor boundaries is refined consistently with the update of these data structures via MPI calls. The parallel adaption scheme has been applied to the solution of an unsteady, three-dimensional, nonlinear, groundwater flow problem. Timings indicate efficiency of the grid refinement process relative to the flow solver calculations.
In high level synthesis for real-time digital signal processing (DSP) architectures using heterogeneous functional units (FUs), an important problem is how to assign a proper FU type to each operation of a DSP applica...
详细信息
ISBN:
(纸本)0769521320
In high level synthesis for real-time digital signal processing (DSP) architectures using heterogeneous functional units (FUs), an important problem is how to assign a proper FU type to each operation of a DSP application and generate a schedule in such a way that all requirements can be met and the total cost can be minimized. In this paper, we propose a two-phase approach to solve this problem. In the first phase, we solve heterogeneous assignment problem, i.e., how to assign a proper FU type to a DSP application such that the total cost can be minimized while the timing constraint is satisfied. In the second phase, based on the assignments obtained from the first phase, we propose a minimum resource scheduling algorithm to generate a schedule and a feasible configuration that uses as little resource as possible. We prove heterogeneous assignment problem is NP-complete and propose several algorithms to solve it. The experiments show that Algorithm DFG_Assign_Repeat is the best that gives a reduction of 25.7% on total cost compared with the previous work.
We present in this paper a novel method to predict application runtimes on backfilling parallel systems. The method is based on mining historical data to obtain important parameters. These parameters are then applied ...
详细信息
ISBN:
(纸本)9780769539393
We present in this paper a novel method to predict application runtimes on backfilling parallel systems. The method is based on mining historical data to obtain important parameters. These parameters are then applied to predict the runtime of future applications. It has been shown in previous works [4], [21] that both underestimate and inaccuracy in prediction have adverse impacts on scheduling performance of backfilling systems. In our study, we try to reduce the number of jobs that are underestimated and reduce the prediction error as much as possible. Comparing with other predictors, experimental results show that our predictor is upto 25% better with respect to the problem of underestimate. Moreover, using the metric proposed in [21] for the accuracy, our predictor improves upto 32%.
In Internet of Thing (ToT) system based on edge intelligence, pay-for-use demand has become a consensus, and the pricing of ToT services has attracted the attention of academia and industry. The pricing method based o...
详细信息
Data availability is a key aspect of modern distributed systems. We discuss an extension of coordination languages based on tuple spaces with programming abstractions for sharing data and guaranteeing availability wit...
详细信息
ISBN:
(纸本)9781538649756
Data availability is a key aspect of modern distributed systems. We discuss an extension of coordination languages based on tuple spaces with programming abstractions for sharing data and guaranteeing availability with different consistency guarantees. Data can be spread over the system according to user-specified replica placement strategies and user-specified consistency requirements. The framework takes care then of low-level management of the replicas, so that the programmer can just focus on the business logic of the application. We advocate that the proposed programming primitives are beneficial for data-oriented applications where different kinds of data may have different needs in terms of availability and consistency.
Various massively parallel computers have emerged in recent years. Each of then have some distinct architecture properties that challenge the computer scientist to develop algorithms and software appropriate to that s...
详细信息
ISBN:
(纸本)0818656026
Various massively parallel computers have emerged in recent years. Each of then have some distinct architecture properties that challenge the computer scientist to develop algorithms and software appropriate to that specified architecture. This is at least as difficult a problem as software reuse in the sequential computer case. One approach to addressing this problem is to design parallel computation models which abstract the architecture details into several generic parameters, which we call resource metrics. Typical resource metrics include the number of processors, communication latency, synchronization, bandwidth, block transfer capability, memory access method, and network topology hierarchy. We review the various parallel and distributed computation models and compare the different resource metrics chosen by different computation models.
The architecture of multiplicative type-II Cellular Neural Networks with Fuzzy (FCNN) is proposed, which is with local connectivity advantageous suitable implemented for VLSI. The fuzzy AND and fuzzy OR morphological ...
详细信息
ISBN:
(纸本)9781424411092
The architecture of multiplicative type-II Cellular Neural Networks with Fuzzy (FCNN) is proposed, which is with local connectivity advantageous suitable implemented for VLSI. The fuzzy AND and fuzzy OR morphological operators in the Type H FCNN can be operated by Min and Max operations for the feel forward template from neighbouring cell of input image in dynamic parallelprocessing. Base on the proposed FCNN structure that the neuron, Min/Max, analog multiplier, transfer and control circuits are adopted and verified by the HSPICE. The desired experimental chip of 3 x 3 multiplicative type-H FCNN is fabricated using TSMC 0.35 mu m. 2P4M CMOS technology. The measurement results have successfully verified the correct function of FCNN. There have a great potential in the VLSI implementation of neural network systems for binary and gray-level patterns in image-processing for consume electronic applications.(1).
暂无评论