This paper analyzes the performance of two parallel algorithms for solving the linear-quadratic optimal control problem arising in discrete-time periodic linear systems. The algorithms perform a sequence of orthogonal...
详细信息
This paper analyzes the performance of two parallel algorithms for solving the linear-quadratic optimal control problem arising in discrete-time periodic linear systems. The algorithms perform a sequence of orthogonal reordering transformations on formal matrix products associated with the periodic linear system and then employ the so-called matrix disk function to solve the resulting discrete-time periodic algebraic Riccati equations needed to determine the optimal periodic feedback. We parallelize these solvers using two different approaches, based on a coarse-grain and a medium-grain distribution of the computational load. The experimental results report the high performance and scalability of the parallel algorithms on a Beowulf cluster. (C) 2002 Elsevier Science (USA).
Binary addition and multiplication problems are very important as their time dominates computation time of any scientific or engineering problem. Simple algorithms are presented for these 2 problems which take only O...
详细信息
Binary addition and multiplication problems are very important as their time dominates computation time of any scientific or engineering problem. Simple algorithms are presented for these 2 problems which take only O(1) time and O(log n) time on a linear PARBS and n x 2n-PARBS respectively, in which each processor has only a constant number of gates and registers. It is believed that these algorithms could be an efficient design for the implementation of an adder and multiplier circuit in a single VLSI chip.
Pichat and Bohlender studied an algorithm for the rounding exact summation of floating point numbers which can be executed on any floating point arithmetic unit. We propose parallel versions of this algorithm, namely ...
详细信息
Pichat and Bohlender studied an algorithm for the rounding exact summation of floating point numbers which can be executed on any floating point arithmetic unit. We propose parallel versions of this algorithm, namely a pipeline version, an algorithm similar to the exchange methods for sorting and a tree-like algorithm, associating a tree to the sum. For all these algorithms we discuss the properties, a multiprocessor architecture should have for an efficient implementation of an algorithm without restricting us to a special architecture.
In this paper we show structural and algorithmic properties on the class of quasi-threshold graphs, or QT-graphs for short, and prove necessary and sufficient conditions for a QT-graph to be Hamiltonian. Based on thes...
详细信息
In this paper we show structural and algorithmic properties on the class of quasi-threshold graphs, or QT-graphs for short, and prove necessary and sufficient conditions for a QT-graph to be Hamiltonian. Based on these properties and conditions, we construct an efficient parallel algorithm for finding a Hamiltonian cycle in a QT-graph;for an input graph on n vertices and in edges, our algorithm takes O(log n) time and requires O(n + m) processors on the CREW PRAM model. In addition, we show that the problem of recognizing whether a QT-graph is a Hamiltonian graph and the problem of computing the Hamiltonian completion number of a nonHamiltonian QT-graph can also be solved in O(log n) time with O(n + in) processors. Our algorithms rely on O(log n)-time parallel algorithms, which we develop here, for constructing tree representations of a QT-graph;we show that a QT-graph G has a unique tree representation, that is, a tree structure which meets the structural properties of G. We also present parallel algorithms for other optimization problems on QT-graphs which run in O(log n) time using a linear number of processors. (C) 2003 Elsevier Inc. All rights reserved.
This paper proposes a parallel algorithm for robot path planning on a linear array with a reconfigurable pipelined bus system (LARPBS) through the construction of a Voronoi diagram on a binary image of the workspace. ...
详细信息
This paper proposes a parallel algorithm for robot path planning on a linear array with a reconfigurable pipelined bus system (LARPBS) through the construction of a Voronoi diagram on a binary image of the workspace. The algorithm is based on a d(4) distance metric, and it does not incur any additional time or processor requirements compared with those of a previously reported proposal (Tzionas et al., 1997). This paper recommends the same model as the simpler VLSI architecture for the problem in question.
Processing of logical data always requires special software and hardware tools. This is attributable to the specific features of the mathematical apparatus of the algebra of logical functions. Organization of parallel...
详细信息
Processing of logical data always requires special software and hardware tools. This is attributable to the specific features of the mathematical apparatus of the algebra of logical functions. Organization of parallel logical computation on the basis of the symbolic mathematical apparatus leads to complex logic programs. A different approach is proposed in this article. It is based on the matrix apparatus. Its use enables us to synthesize parallel and structurally homogeneous algorithms for the evaluation of directional logical derivatives of multivalued logic functions and implement their evaluation using standard matrixalgebra software or homogeneous computing systems. Homogeneous computing systems substantially accelerate the processing speed and can be built using VLSI technology.
The operation graphs of the proposed algorithms have the same configuration as the graphs of fast algorithms used in digital signal processing. This result makes it possible to use well-tried standard procedures of digital signal processing, which involve mapping of algorithms into homogeneous computing structures and hardware-software architectures.
One of the most ambitious trends in current biomedical research is the large-scale genomic sequencing of patients. Novel high-throughput (or next-generation) sequencing technologies have redefined the way genome seque...
详细信息
One of the most ambitious trends in current biomedical research is the large-scale genomic sequencing of patients. Novel high-throughput (or next-generation) sequencing technologies have redefined the way genome sequencing is performed. They are able to produce millions of short sequences (reads) in a single experiment, and with a much lower cost than previously possible. Due to this massive amount of data, efficient algorithms for mapping these sequences to a reference genome are in great demand, and recently, there has been ample work for publishing such algorithms. One important feature of these algorithms is the support of multithreaded parallel computing in order to speedup the mapping process. In this paper, we design parallel algorithms, which make use of the message-passing parallelism model, to address this problem efficiently. The proposed algorithms also take into consideration the probability scores assigned to each base for occurring in a specific position of a sequence. In particular, we present parallel algorithms for mapping short degenerate and weighted DNA sequences to a reference genome.
Multi-level network optimization (MLNO) problems arise in many contexts such as telecommunication, transportation, or electric power systems. This paper is mainly concerned with parallel implementations of the classic...
详细信息
Multi-level network optimization (MLNO) problems arise in many contexts such as telecommunication, transportation, or electric power systems. This paper is mainly concerned with parallel implementations of the classical branch-and-bound algorithm for multi-level network design. A model for such a problem is presented and formulated as a mixed-integer program. The formulation is appealing because it integrates in the same model aspects of discrete facility location, topological network design, and dimensioning. We propose implementations that are suitable for multiple instruction stream, multiple data stream (MIMD) parallel computation systems. Thus, the implementations are very convenient for use in networks of workstations, which nowadays has become so popular. We have tested two versions of the branch-and-bound algorithm as well as different load balancing strategies. The results are very encouraging indicating a gain over sequential computations in terms of execution time.
The ability to model the temporal dimension is essential to many applications. Furthermore, the rate of increase in database size and response time requirements has outpaced advancements in processor and mass storage ...
详细信息
The ability to model the temporal dimension is essential to many applications. Furthermore, the rate of increase in database size and response time requirements has outpaced advancements in processor and mass storage technology, leading to the need for parallel temporal database management systems. In this paper, we introduce a variety of parallel temporal aggregation algorithms for a shared-nothing architecture based on the sequential Aggregation Tree algorithm. Via an empirical study, we found that the number of processing nodes, the partitioning of the data, the placement of results, and the degree of data reduction effected by the aggregation impacted the performance of the algorithms. For distributed results placement, we discovered that Time Division Merge was the obvious choice. For centralized results and high data reduction, Pairwise Merge was preferred regardless of the number of processing nodes, but for low data reduction, it only performed well up to 32 nodes. This led us to a centralized variant of Time Division Merge which was best for larger configurations having low data reduction.
Given a set of n intervals representing an interval graph, the problem of finding a maximum matching between pairs of disjoint (nonintersecting) intervals has been considered in the sequential model. In this paper we ...
详细信息
Given a set of n intervals representing an interval graph, the problem of finding a maximum matching between pairs of disjoint (nonintersecting) intervals has been considered in the sequential model. In this paper we present parallel algorithms for computing maximum cardinality matchings among pairs of disjoint intervals in interval graphs in the EREW PRAM and hypercube models. For the general case of the problem, our algorithms compute a maximum matching in O (log(3) n) time using O (nl log(2) n) processors on the EREW PRAM and using n processors on the hypercubes. For the case of proper interval graphs, our algorithm runs in O (log n) time using O (n) processors if the input intervals are not given already sorted and using O (n/log n) processors otherwise, on the EREW PRAM. On n-processor hypercubes, our algorithm for the proper interval case takes O (log n log log n) time for unsorted input and O (log n) time for sorted input. Our parallel results also lead to optimal sequential algorithms for computing maximum matchings among disjoint intervals. In addition, we present an improved parallel algorithm for maximum matching between overlapping intervals in proper interval graphs.
暂无评论