The performances of some popular routing algorithms used for multicast communications are extensively simulated and analyzed in comparative terms, assuming the Bidirectional Manhattan Street Network as a valued test-b...
详细信息
ISBN:
(纸本)0819428604
The performances of some popular routing algorithms used for multicast communications are extensively simulated and analyzed in comparative terms, assuming the Bidirectional Manhattan Street Network as a valued test-bed. Multicast traffic is inserted over a substrate of point-to-point communications, contributing to increase the value of some cost parameters, suitably defined to express the routing efficiency. Both the cases of uncorrelated and bursty traffic are considered. The analysis confirms that the superiority of an algorithm over the others cannot be claimed since it depends on the input variables and on the criterion used for performance evaluation. Anyway conclusions can be drawn under specific assumptions, and they are usable in the framework of the network design.
In a dedicated, mixed-machine, heterogeneous computing (HC) system, an application program may be decomposed into subtasks, then each subtask assigned to the machine where it is best suited for execution. Data relocat...
详细信息
In a dedicated, mixed-machine, heterogeneous computing (HC) system, an application program may be decomposed into subtasks, then each subtask assigned to the machine where it is best suited for execution. Data relocation is defined as selecting the sources for needed data items. It is assumed that multiple independent subtasks of an application program can be executed concurrently on different machines whenever possible. A theoretical stochastic model for HC is proposed, in which the computation times of subtasks and communication times for intermachine data transfers can be random variables. The optimization problem for finding the optimal matching, scheduling, and data relocation schemes to minimize the total execution time of an application program is defined based on this stochastic HC model. The global optimization criterion and search space for the above optimization problem are described. It is validated that a greedy algorithm-based approach can establish a local optimization criterion for developing data relocation heuristics. The validation is provided by a theoretical proof based on a set of common assumptions about the underlying HC system and application program. The local optimization criterion established by the greedy approach, coupled with the search space defined for choosing valid data relocation schemes, can help developers of future practical data relocation heuristics.
The QR orthogonal decomposition method is well established for solving linear systems and both sequential and parallel implementations have been proposed. In this paper, the QZ orthogonal decomposition method is propo...
详细信息
The QR orthogonal decomposition method is well established for solving linear systems and both sequential and parallel implementations have been proposed. In this paper, the QZ orthogonal decomposition method is proposed in which 2 elements are eliminated simultaneously. Consequently, the theoretical analyses presented show that the QZ method is faster than the QR method which is confirmed by the numerical results.
Let S(v) be a function defined on the vertices v of the infinite binary tree. One algorithm to seek large positive values of S is the Metropolis-type Markov chain (X-n) defined by P(Xn+1 = w \ X-n = v) = 1/3 e(b(S(w)-...
详细信息
Let S(v) be a function defined on the vertices v of the infinite binary tree. One algorithm to seek large positive values of S is the Metropolis-type Markov chain (X-n) defined by P(Xn+1 = w \ X-n = v) = 1/3 e(b(S(w)-S(v)))/1 + e(b(S(w)-S(v))) for each neighbor w of v, where b is a parameter ("1/temperature") which the user can choose. We introduce and motivate study of this algorithm under a probability model for the objective function S, in which S is a "tree-indexed simple random walk," that is, the increments xi(e) = S(w) - S(v) along parent-child edges e = (v, w) are independent and P(xi = 1) = p, P(xi = -1) = 1 - p. This algorithm has a "speed" r(p, b) = lim(n)n(-1) ES(X-n). We study the speed via a mixture of rigorous arguments, nonrigorous arguments, and Monte Carlo simulations, and compare with a deterministic greedy algorithm which permits rigorous analysis. Formalizing the nonrigorous arguments presents a challenging problem. Mathematically, the subject is in part analogous to recent work of Lyons et al. [19], [20] on the speed on random walk on Galton-Watson trees. A key feature of the model is the existence of a critical point p(crit) below which the problem is infeasible;we study the behavior of algorithms as p down arrow p(crit). This provides a novel analogy between optimization algorithms and statistical physics.
We define the base polytope B(P, g) of partially ordered set P and a supermodular function g on the ideals of P as the convex hull of the incidence vectors of all linear extensions of P. This new class of polytopes co...
详细信息
We define the base polytope B(P, g) of partially ordered set P and a supermodular function g on the ideals of P as the convex hull of the incidence vectors of all linear extensions of P. This new class of polytopes contains, among others, the base polytopes of supermodular systems and permutahedra as special cases. After introducing the notion of compatibility for g, we give a complete linear description of B(P, g) for series-parallel posets and compatible functions g. In addition, we describe a greedy-type procedure which exhibits Sidney's job sequencing algorithm to minimize the total weighted completion time as a natural extension of the matroidal greedy algorithm from sets to posets. (C) 1998 The Mathematical Programing Society, Inc. Published by Elsevier Science B.V.
Active contour models for object tracking are investigated. The unbiased active contour algorithm, a model using edge feature based tracking, modifies the energy functions to correct the bias towards equal-distanced s...
详细信息
Active contour models for object tracking are investigated. The unbiased active contour algorithm, a model using edge feature based tracking, modifies the energy functions to correct the bias towards equal-distanced snaxels and small curvature, and eliminates the need to treat corners specially. Comparison with the William and Shah (1992) model shows substantial improvements. (C) 1998 Published by Elsevier Science B.V. All rights reserved.
A new fast vector quantiser (VQ) is proposed which reduces computational complexity by searching the intersection of starch regions defined by triangular inequalities for multiple anchors. A greedy algorithm which sel...
详细信息
A new fast vector quantiser (VQ) is proposed which reduces computational complexity by searching the intersection of starch regions defined by triangular inequalities for multiple anchors. A greedy algorithm which selects an anchor set is also proposed. The proposed VQ reduces the number of multiply operations by 24.4% compared to Li's VQ [4] while preserving the same encoding quality as the full-search VQ.
For a given congruence condition, we try to find a subset-sum-distinct sequence such that the sum of the reciprocals of its terms is maximal. We show that the greedy algorithm generates an extremal sequence infinitely...
详细信息
For a given congruence condition, we try to find a subset-sum-distinct sequence such that the sum of the reciprocals of its terms is maximal. We show that the greedy algorithm generates an extremal sequence infinitely often but also fails to generate an extremal sequence infinitely often. We also provide two criteria for determining whether the sequence provided by the greedy algorithm is maximal or not. (C) 1998 Elsevier science B.V. All rights reserved.
Let X = (X-1,..., X-n) be a random binary vector, with a known joint distribution P. It is necessary to inspect the coordinates sequentially in order to determine if X-i = 0 for every i, i = 1,..., n. We find bounds f...
详细信息
Let X = (X-1,..., X-n) be a random binary vector, with a known joint distribution P. It is necessary to inspect the coordinates sequentially in order to determine if X-i = 0 for every i, i = 1,..., n. We find bounds for the ratio of the expected number of coordinates inspected using optimal and greedy searching policies.
A set D is called greedy if the greedy algorithm solves the problem max {cx \x is an element of D} for any c such that c(1) greater than or equal to ... greater than or equal to c(n) greater than or equal to 0. In pre...
详细信息
A set D is called greedy if the greedy algorithm solves the problem max {cx \x is an element of D} for any c such that c(1) greater than or equal to ... greater than or equal to c(n) greater than or equal to 0. In previous articles we described the class of all greedy convex polyhedra in R-n using the partial order technique. Here we generalize the problem and extend the class of greedy feasible sets using the so-called generalized canonical order. These results are spread onto some related problems (characterization of greedy matrices and functions). (C) 1997 Published by Elsevier Science B.V.
暂无评论