A model for parallel and distributed programs, the dynamic process graph (DPG), is investigated under graph-theoretic and complexity aspects. Such graphs embed constructors for parallel programs, synchronization mechanisms as well as conditional branches. They are capable of representing all possible executions of a parallel or distributed program in a very compact way. The size of this representation can be as small as logarithmic with respect to the size of any execution of the program.
In a preceding paper [A. Jakoby, et al., Scheduling dynamic graphs, in: Proc. 16th Symposium on Theoretical Aspects in Computer Science STACS'99, LNCS, vol. 1563, Springer, 1999, pp. 383–392] we have analysed the expressive power of the general model and various variants of it. We have considered the scheduling problem for DPGs given enough parallelism taking into account communication delays between processors when exchanging data. Given a DPG the question arises whether it can be executed (that means whether the corresponding parallel program has been specified correctly), and what is its minimum schedule length.
In this paper we study a subclass of dynamic process graphs called
-output DPGs, which are appropriate in many situations, and investigate their expressive power. In a previous paper we have shown that the problem to determine the minimum schedule length is still intractable for this subclass, namely this problem is
-complete as is the general case. Here we will investigate structural properties of the executions of such graphs. A natural graph-theoretic conjecture that executions must always split into components that are isomorphic to subgraphs turns out to be wrong. We are able to prove a weaker property. This implies a quadratic upper bound on the schedule length that may be necessary in the worst case, in contrast to the general case, where the optimal schedule length may be exponential with respect to the size of the representing DPG. Making this bound constructive, we obtain an approximation to a
-complete problem. Computing such a schedule and then executing the program can be done on a parallel machine in polynomial time in a highly distributive fashion. 相似文献
In this paper a tripartite qualitative design combining abservation, stimulated recall and interview is presented and discussed. This three-step-design makes it possible to get insight into the interaction of internal and external processes when solving mathematical tasks. The data analysis depends on the research question and the methodological approach. In the light of two research projects in mathematics education two different methods of data analysis are presented and methodologically reflected. 相似文献
The title compound, {(C9H14N)4[Pb3I10]}n, crystallizes as an organic–inorganic hybrid. As such, the structure consists of a two‐dimensional inorganic layer of [Pb3I10]n4n− ions extending along [100]. The asymmetric unit contains two independent Pb atoms, viz. one in a general position and the other on an inversion centre. Each Pb atom is octahedrally coordinated by six iodide ions and exhibits both face‐ and corner‐sharing with adjacent atoms in the inorganic layer. These anionic layers alternate with 3‐phenylpropylammonium cations, which hydrogen bond to the iodides. Simple face‐to‐edge σ–π stacking interactions are observed between the aromatic rings that stabilize the overall three‐dimensional structure. This net structure has only been observed five times previously. 相似文献
This article introduces and analyzes a p-version FEM for variational inequalities resulting from obstacle problems for some quasi-linear elliptic partial differential
operators. We approximate the solution by controlling the obstacle condition in images of the Gauss–Lobatto points. We show
existence and uniqueness for the discrete solution up from the p-version for the obstacle problem. We prove the convergence of up towards the solution with respect to the energy norm, and assuming some additional regularity for the solution we derive
an a priori error estimate. In numerical experiments the p-version turns out to be superior to the h-version concerning the convergence rate and the number of unknowns needed to achieve a certain exactness of the approximation. 相似文献
The paper deals with complementarity problems CP(F), where the underlying functionF is assumed to be locally Lipschitzian. Based on a special equivalent reformulation of CP(F) as a system of equationsφ(x)=0 or as the problem of minimizing the merit functionΘ=1/2∥Φ∥
22
, we extend results which hold for sufficiently smooth functionsF to the nonsmooth case.
In particular, ifF is monotone in a neighbourhood ofx, it is proved that 0 εδθ(x) is necessary and sufficient forx to be a solution of CP(F). Moreover, for monotone functionsF, a simple derivative-free algorithm that reducesΘ is shown to possess global convergence properties. Finally, the local behaviour of a generalized Newton method is analyzed.
To this end, the result by Mifflin that the composition of semismooth functions is again semismooth is extended top-order semismooth functions. Under a suitable regularity condition and ifF isp-order semismooth the generalized Newton method is shown to be locally well defined and superlinearly convergent with the
order of 1+p. 相似文献
The theory of nonequilibrium potentials or quasipotentials is a physically motivated approach to small random perturbations of dynamical systems, leading to exponential estimates of invariant probabilities and mean first exit times. In the present article we develop the mathematical foundation of this theory for discrete-time systems, following and extending the work of Freidlin and Wentzell, and Kifer. We discuss strategies for calculating and estimating quasipotentials and show their application to one-dimensionalS-unimodal maps. The method proves to be especially suited for describing the noise scaling behavior of invariant probabilities, e.g., for the map occurring as the limit of the Feigenbaum period-doubling sequence. We show that the method allows statements about the scaling behavior in the case of localized noise, too, which does not originally lie within the scope of the quasipotential formalism. 相似文献
The projected areas of non-spherical particles do not represent an unambiguous particle characteristic. Depending on the orientation towards a constant observational direction, different projected areas result. The spectrum of all projected area values of a particle, if determined representatively, gives the probability with which a certain value is obtained by a single measurement. In this work, the frequency distributions of different examples of test objects were both calculated and measured. The objects were a cube, a rectangular parallelepiped and also three model agglomerates consisting of spheres of the same size. Instead of just one projected area, during each measuring procedure three projected areas from three orthogonal directions can be obtained. A mean value is then calculated to reduce the ambiguity of the particle characteristic and enhance the resolution. A suitable measurement set-up is introduced. The results of calculation and measurement are compared for observation from just one direction and also simultaneous observation from three directions. The frequency distributions of the equivalent diameters of the particle projected areas show a characteristic trend of the total curve with remarkable properties. The simultaneous measurement of three values from mutually orthogonal directions and their mean value calculation result in a much narrower distribution. In this case, a non-sphericity factor can additionally be calculated, whose frequency distribution contains information in a characteristic manner about the degree to which the particle shape differs from a sphere. 相似文献