Scheduling Malleable Tasks on Parallel Processors to Minimize the Makespan

The problem of optimal scheduling n tasks in a parallel processor system is studied. The tasks are malleable, i.e., a task may be executed by several processors simultaneously and the processing speed of a task is a nonlinear function of the number of processors allocated to it. The total number of processors is m and it is an upper bound on the number of processors that can be used by all the tasks simultaneously. It is assumed that the number of processors is sufficient to process all the tasks simultaneously, i.e. n≤m. The objective is to find a task schedule and a processor allocation such that the overall task completion time, i.e. the makespan, is minimized. The problem is motivated by real-life applications of parallel computer systems in scientific computing of highly parallelizable tasks. An O(n) algorithm is presented to solve this problem when all the processing speed functions are convex. If these functions are all concave and the number of tasks is a constant, the problem can be solved in polynomial time. A relaxed problem, in which the number of processors allocated to each task is not required to be integer, can be solved in O(nmax {m,nlog ² m}) time. It is proved that the minimum makespan values for the original and relaxed problems coincide. For n=2 or n=3, an optimal solution for the relaxed problem can be converted into an optimal solution for the original problem in a constant time.     

An extension of Karmarkar's algorithm for solving a system of linear homogeneous equations on the simplex

We present an extension of Karmarkar's algorithm for solving a system of linear homogeneous equations on the simplex. It is shown that in at most O(nL) steps, the algorithm produces a feasible point or proves that the problem has no solution. The complexity is O(n ² m ² L) arithmetic operations. The algorithm is endowed with two new powerful stopping criteria.     

A weighted independent even factor algorithm

An even factor in a digraph is a vertex-disjoint collection of directed cycles of even length and directed paths. An even factor is called independent if it satisfies a certain matroid constraint. The problem of finding an independent even factor of maximum size is a common generalization of the nonbipartite matching and matroid intersection problems. In this paper, we present a primal-dual algorithm for the weighted independent even factor problem in odd-cycle-symmetric weighted digraphs. Cunningham and Geelen have shown that this problem is solvable via valuated matroid intersection. Their method yields a combinatorial algorithm running in O(n ³ γ +? n ⁶ m) time, where n and m are the number of vertices and edges, respectively, and γ is the time for an independence test. In contrast, combining the weighted even factor and independent even factor algorithms, our algorithm works more directly and runs in O(n ⁴ γ?+?n ⁵) time. The algorithm is fully combinatorial, and thus provides a new dual integrality theorem which commonly extends the total dual integrality theorems for matching and matroid intersection.     

An O(K·n4) algorithm for finding the k best cuts in a network

An algorithm for finding the K best cuts in a network is presented. Using a branch technique introduced by Lawler [4] we reduce the problem to K computations of 2nd best cuts. The latter problem can be solved by an O(n⁴) algorithm yielding an overall complexity of O(K·n⁴) for the presented algorithm.     

The Hamiltonian problem on distance-hereditary graphs

In this paper, we first present an O(n+m)-time sequential algorithm to solve the Hamiltonian problem on a distance-hereditary graph G, where n and m are the number of vertices and edges of G, respectively. This algorithm is faster than the previous best known algorithm for the problem which takes O(n²) time. We also give an efficient parallel implementation of our sequential algorithm. Moreover, if G is represented by its decomposition tree form, the problem can be solved optimally in O(logn) time using O((n+m)/logn) processors on an EREW PRAM.     

New polynomial-time algorithms for Camion bases

Let M be a finite set of vectors in Rⁿ of cardinality m and H(M)={{x∈Rⁿ:c^Tx=0}:c∈M} the central hyperplane arrangement represented by M. An independent subset of M of cardinality n is called a Camion basis, if it determines a simplex region in the arrangement H(M). In this paper, we first present a new characterization of Camion bases, in the case where M is the column set of the node-edge incidence matrix (without one row) of a given connected digraph. Then, a general characterization of Camion bases as well as a recognition procedure which runs in O(n²m) are given. Finally, an algorithm which finds a Camion basis is presented. For certain classes of matrices, including totally unimodular matrices, it is proven to run in polynomial time and faster than the algorithm due to Fonlupt and Raco.     

The tricriterion shortest path problem with at least two bottleneck objective functions

The focus of this paper is on the tricriterion shortest path problem where two objective functions are of the bottleneck type, for example MinMax or MaxMin. The third objective function may be of the same kind or we may consider, for example, MinSum or MaxProd. Let p(n) be the complexity of a classical single objective algorithm responsible for this third function, where n is the number of nodes and m be the number of arcs of the graph. An O(m²p(n)) algorithm is presented that can generate the minimal complete set of Pareto-optimal solutions. Finding the maximal complete set is also possible. Optimality proofs are given and extensions for several special cases are presented. Computational experience for a set of randomly generated problems is reported.     

A fast algorithm for solving systems of linear equations with two variables per equation

We present a fast algorithm for solving m X n systems of linear equations Ax = c with at most two variables per equation. The algorithm makes use of a linear-time algorithm for constructing a spanning forest of an undirected graph, and it requires 5m + 2n – 2 arithmetic operations in the worst case.     

An algorithm for finding hamilton cycles in random directed graphs

We describe a polynomial (O(n^1.5)) time algorithm DHAM for finding hamilton cycles in digraphs. For digraphs chosen uniformly at random from the set of digraphs with vertex set {1, 2, …, n} and m = m(n) edges the limiting probability (as n → ∞) that DHAM finds a hamilton cycle equals the limiting probability that the digraph is hamiltonian. Some applications to random “travelling salesman problems” are discussed.     

David G. Luenberger,Linear and nonlinear programming, 2nd edition,Addison-Wesley,Reading (1984), xv + 492 pages, $35.50

This paper considers a stochastic version of the linear continuous type knapsack problem in which the cost coefficients are random variables. The problem is to find an optimal solution and an optimal probability level of the chance constraint. This problem P₀ is first transformed into a deterministic equivalent problem P. Then a subproblem with a positive parameter is introduced and a close relation between P and its subproblem is shown. Further, an auxiliary problem of the subproblem is introduced and a direct relation between P and the auxiliary problem is derived through a relation connecting the subproblem and its auxiliary problem. Fully utilizing these relations, an efficient algorithm is proposed that finds an optimal solution of P in at most O(n⁴) computational time where n is the number of decision variables. Finally, further research problems are discussed.     

A direct method to solve block banded block Toeplitz systems with non-banded Toeplitz blocks

A fast solution algorithm is proposed for solving block banded block Toeplitz systems with non-banded Toeplitz blocks. The algorithm constructs the circulant transformation of a given Toeplitz system and then by means of the Sherman-Morrison-Woodbury formula transforms its inverse to an inverse of the original matrix. The block circulant matrix with Toeplitz blocks is converted to a block diagonal matrix with Toeplitz blocks, and the resulting Toeplitz systems are solved by means of a fast Toeplitz solver.The computational complexity in the case one uses fast Toeplitz solvers is equal to ξ(m,n,k)=O(mn³)+O(k³n³) flops, there are m block rows and m block columns in the matrix, n is the order of blocks, 2k+1 is the bandwidth. The validity of the approach is illustrated by numerical experiments.     

Cycle-free partial orders and chordal comparability graphs

This paper studies a number of problems on cycle-free partial orders and chordal comparability graphs. The dimension of a cycle-free partial order is shown to be at most 4. A linear time algorithm is presented for determining whether a chordal directed graph is transitive, which yields an O(n ²) algorithm for recognizing chordal comparability graphs. An algorithm is presented for determining whether the transitive closure of a digraph is a cycle-free partial order in O(n+m _t)time, where m _tis the number of edges in the transitive closure.     

A simple version of Karzanov's blocking flow algorithm

Dinic has shown that the classic maximum flow problem on a graph of n vertices and m edges can be reduced to a sequence of at most n ? 1 so-called ‘blocking flow’ problems on acyclic graphs. For dense graphs, the best time bound known for the blocking flow problems is O(n²). Karzanov devised the first O(n²)-time blocking flow algorithm, which unfortunately is rather complicated. Later Malhotra, Kumar and Maheshwari devise another O(n²)-time algorithm, which is conceptually very simple but has some other drawbacks. In this paper we propose a simplification of Karzanov's algorithm that is easier to implement than Malhotra, Kumar and Maheshwari's method.     

On spline function approximations to the solution of volterra integral equations of the first kind

A procedure, using spline functions of degreem, for the solution of linear Volterra integral equations of the first kind is presented. The method produces an approximate solution of classC ^m-1, is order (m+1) and is shown to be numerically stable form≦4.     

PCR algorithm for parallel computing the solution of the general restricted linear equations

In this paper we present a parallel algorithm for parallel computing the solution of the general restricted linear equations Ax=b,x∈T, where T is a subspace of ? ⁿ and b∈AT. By this algorithm the solution x=A _T,S ⁽²⁾ b is obtained in n(log?₂ m+log?₂(n?s+1)+7)+log?₂ m+1 steps with P=mn processors when m≥2(n?1) and with P=2n(n?1) processors otherwise.     

A graph-oriented approach for the minimization of the number of late jobs for the parallel machines scheduling problem

An O(n²) algorithm is presented for the n jobs m parallel machines problem with identical processing times. Due dates for each job are given and the objective is the minimization of the number of late jobs. Preemption is permitted. The problem can be formulated as a maximum flow network model. The optimality proof as well as other properties and a complete example are given.     

Efficient Comparison Based String Matching

We study the exact number of symbol comparisons that are required to solve the string matching problem and present a family of efficient algorithms. Unlike previous string matching algorithms, the algorithms in this family do not "forget" results of comparisons, what makes their analysis much simpler. In particular, we give a linear-time algorithm that finds all occurrences of a pattern of length m in a text of length n in [formula] comparisons. The pattern preprocessing takes linear time and makes at most 2m comparisons. This algorithm establishes that, in general, searching for a long pattern is easier than searching for a short one. We also show that any algorithm in the family of the algorithms presented must make at least [formula] symbol comparisons, for m = 2^k − 1 and any integer k ≥ 1.     

An algorithm for a separable integer programming problem with cumulatively bounded variables

An algorithm is presented for obtaining the optimal solution of an integer programming problem in which the nested constraints represent the cumulative bounding of the variables and the objective function is a sum of concave functions of one variables. The algorithm requires O(m log(m)log²(b_m/m)) time, where m is the number of variables and b_m is an upper bound on the sum of the m variables, b_m ⩾m⩾1. It is also demonstrated that a special case of identical concave functions is solvable in O(m). Both results significantly improve the previous bounds for these problems.     

An efficient algorithm to solve connectivity problem on trapezoid graphs

The connectivity problem is a fundamental problem in graph theory. The best known algorithm to solve the connectivity problem on general graphs withn vertices andm edges takesO(K(G)mn ^1.5) time, whereK(G) is the vertex connectivity ofG. In this paper, an efficient algorithm is designed to solve vertex connectivity problem, which takesO(n ²) time andO(n) space for a trapezoid graph.     

An assignment algorithm with applications to integrated circuit layout

The 2-terminal one-to-any problem, which arises in the design of layout systems, is the problem of assigning wach one of n terminals positioned on the upper row of a channel (called entry terminals) to one of m terminals positioned on the lower row (called exit terminals) so that the resulting channel routing problem has minimum density. An optimal solution to this problem is known [1]. In this paper we consider a natural generalization, the 2-color one-to-any problem, in which we have two types of entry terminals, red and blue ones, and exit terminals can be assigned to either type of entry terminal. Red and blue nets created by our algorithm are allowed to run on top of each other in the routing, and the density is defined as the larger of the red density and the blue density. Its minimization is an interesting combinatorial problem. We show how to compute the best achievable density in O(n + m) time, and an assignment achieving this density in O((n + m)log(n + m)) time.