In this paper a systematic methodology for designing parallel-prefix modulo 2n - 1 adders, for every n, is introduced. The resulting modulo 2n - 1 adders feature minimum logical dept. and bounded fan-out loading. Addi...
详细信息
In this paper a systematic methodology for designing parallel-prefix modulo 2n - 1 adders, for every n, is introduced. The resulting modulo 2n - 1 adders feature minimum logical dept. and bounded fan-out loading. Additionally, an optimization technique is proposed, which aims to the reduction of redundant operators that appear on the parallel-prefix carry computation trees. Performance data reveal that the reduced structures achieve area × time complexity reduction of up to 46% when compared to previously reported designs.
It has recently been shown that accumulators can be used efficiently for test pattern generation as well as for test response compaction. In this paper we present a BIST scheme for accumulators where the accumulator i...
详细信息
It has recently been shown that accumulators can be used efficiently for test pattern generation as well as for test response compaction. In this paper we present a BIST scheme for accumulators where the accumulator is simultaneously used as a test pattern generator and a response compactor during its own testing. We also show that the proposed BIST scheme is especially suitable for accumulator, adder and multiplier-accumulator RNS channels leading to minimal hardware overhead and short test sequences.
A program to automate the finite element method (FEM) using various soft-computing techniques is presented. The overall program is discussed, and the implementations of three specific sub-problems (mesh placement, nod...
详细信息
The problem of cooperatively performing a collection of tasks in a decentralized setting where the computing medium is subject to adversarial perturbations is one of the fundamental problems in distributed computing. ...
详细信息
ISBN:
(纸本)1581136048
The problem of cooperatively performing a collection of tasks in a decentralized setting where the computing medium is subject to adversarial perturbations is one of the fundamental problems in distributed computing. Such perturbations can be caused by processor failures, unpredictable delays, and communication breakdowns. To develop efficient distributed solutions for computation problems ranging from distributed search such as SETI to parallel simulation and multi-agent collaboration, it is important to understand efficiency trade-offs characterizing the ability of p processors to cooperate on t-tasks in the presence of adversity. This paper surveys recent results grouped by the following topics: (i) failure-sensitive bounds for distributed cooperation problems for synchronous processors subject to crash failures, (ii) bounds on redundant work for distributed cooperation when individual asynchronous processors may experience prolonged absence of communication, and (in) competitive analysis of cooperative work performed by groups of asynchronous processors, when the groups may be fragmented and merged during the computation. These research results are motivated by the earlier work of the third author with Paris C. Kanellakis at Brown University.
This paper presents a new two-dimensional (2-D) 8×8 discrete cosine/inverse discrete cosine transform (DCT/IDCT) core design using the group distributed arithmetic (GDA) approach. We adopt the way of DA computati...
详细信息
This paper presents a new two-dimensional (2-D) 8×8 discrete cosine/inverse discrete cosine transform (DCT/IDCT) core design using the group distributed arithmetic (GDA) approach. We adopt the way of DA computation and exploit the good features of the cyclic convolution to facilitate an efficient realization of 2-D 8×8 DCT/IDCT core design using shared ROM modules, barrier shifters, and accumulators. To increase the ROM utilization, we re-arrange the content of ROM into several groups in which all the elements in a group will be accessed simultaneously in accumulating the DCT/IDCT outputs. The comparison results with the existing designs show that the proposed design possesses averagely 62.6 % reduction in the delay-area products (ns*Kμm2) based on a 0.35μm CMOS technology.
A system-on-a-chip with reconfigurable hardware such as SRAM-based FPGAs is evidently suitable for several applications on wearable computers. However, interconnect delay of current stateof-the-art FPGAs dominates the...
详细信息
A system-on-a-chip with reconfigurable hardware such as SRAM-based FPGAs is evidently suitable for several applications on wearable computers. However, interconnect delay of current stateof-the-art FPGAs dominates the logic delay with relatively high power consumption. To counterbalance these two problems while maintaining desirable performance, voltage downscaling and fine-grained pipelining have been suggested. In this paper, we propose a pipelining methodology to utilize intra-iteration parallelism of a systolic-like array realized in a multimillion-gate FPGA. This results in not only enhanced throughput but also much higher hardware utilization. Thus, we call it the supersystolic array.
In this paper, we describe an efficient algorithm to select ph onetically balanced scripts for collecti ng a large-scale multilingual speech corpus. It is expected to collect a multilingual speech corpus covering thre...
详细信息
This paper proposes an on-demand distributed clustering algorithm for self-organizing, multihop, mobile packet radio network where mobile nodes are organized into nonoverlapping clusters. These clusters are independen...
详细信息
ISBN:
(纸本)0889863741
This paper proposes an on-demand distributed clustering algorithm for self-organizing, multihop, mobile packet radio network where mobile nodes are organized into nonoverlapping clusters. These clusters are independently controlled and are dynamically reconfigured as nodes may move from one cluster to another cluster. The main advantages of network are to provide spatial reuse of bandwidth due to node clustering and to share or reuse bandwidth in a controlled fashion in each cluster. The proposed algorithm is robust due to the motion, failure, insertion or deletion of nodes. This non periodic algorithm for clusterhead election reduces the computational cost and communication costs.
This paper proposes a simple control method to improve the ignition behavior of cold cathode fluorescent lamp in digital-dimming control. The half-bridge resonant inverter is employed in the presented backlight system...
详细信息
This paper proposes a simple control method to improve the ignition behavior of cold cathode fluorescent lamp in digital-dimming control. The half-bridge resonant inverter is employed in the presented backlight system. To extend the lamp life, we realize a digital-dimming controller with soft-starting technique (DDC-SST) to reduce the high ignition voltage and to eliminate the ignition current spike. Complete analysis and design considerations are discussed in detail in this paper. Simulation and experimental results are close to the theoretical prediction. The overall efficiency of the system achieved at the rated power is over 91%. The ignition voltage is reduced about 30% without any lamp current spike occurred under digital-dimming operation.
The approach described in this article is premised on the idea that drug and alcohol use-related problems are heterogeneously distributed with respect to population and geography, and therefore, are essentially local ...
详细信息
The approach described in this article is premised on the idea that drug and alcohol use-related problems are heterogeneously distributed with respect to population and geography, and therefore, are essentially local problems. More specifically, it is argued that viewing a local community as an interacting set of systems that support or buffer the occurrence of specific substance misuse outcomes, opens up to research two important prospects. The first of these involves creating adequate systems models that can capture the primary community structures and relationships that support public health problems such as alcohol and drug misuse and related outcomes. The second entails rationally testing control strategies that have the potential to moderate or reduce these problems. Understanding and controlling complex dynamic systems models nowadays pervades all scientific disciplines, and it is to research in areas such as biology, ecology, engineering, computer sciences, and mathematics that researchers in the field of addictions must turn to in order to better study the complexity that confronts them as they try to understand and prevent problems resulting from alcohol and drug use and misuse. Here we set out what such a systems-based understanding of alcohol- and drug use-related problems will require and discuss its implications for public policy and prevention programming.
暂无评论