Parallel Simulated Algorithm Research Articles

Design and Benefit Analysis of Distributed Parallel Simulation of Microscopic Traffic System

AbstractThe design of distributed parallel simulation structure and algorithm according to parallel characteristics of traffic network simulation are presented. PVM (Parallel Virtual Machine) serves as parallel computing platform. It puts forwards network-partition algorithm based on number of vehicles and communication means of synchronization to realize load balancing. Parallel benefits of simulation algorithm are assessed by four measures including program run time, speedup, parallel efficiency and expandability. With parallel algorithm, traffic network that is made of 24 intersections is modeled based on different numbers of processors. The results from the simulation demonstrate that simulation speed increases by 3.5 times. Hence, parallel simulation algorithm improves speed and efficiency of traffic network simulation. It provides a foundation for larger scale traffic network simulation.

A parallel genetic simulated annealing hybrid algorithm for task scheduling

In this paper combiend with the advantages of genetic algorithm and simulated annealing, brings forward a parallel genetic simulated annealing hybrid algorithm (PGASHA) and applied to solve task scheduling problem in grid computing. It first generates a new group of individuals through genetic operation such as reproduction, crossover, mutation, etc, and than simulated anneals independently all the generated individuals respectively. When the temperature in the process of cooling no longer falls, the result is the optimal solution on the whole. From the analysis and experiment result, it is concluded that this algorithm is superior to genetic algorithm and simulated annealing.

Parallel discrete event simulation algorithm for manufacturing supply chains

Parallel discrete event simulation (PDES) is concerned with the distributed execution of large-scale system models on multiple processors. It is an enabler in the implementation of the virtual enterprise concept, integrating semi-autonomous models of production cells, factories, or units of a supply chain. The key issue in PDES is to maintain causality relationships between system events, while maximizing parallelism in their execution. Events can be executed conservatively only when it is safe to do so, sacrificing the extent to which potential parallelism of the system can be exploited. Alternatively, they can be processed optimistically without guarantee of correctness, but incurring the overhead of a rollback to an earlier saved state when causality error is detected. The paper proposes a modified optimistic scheme for distributed simulation of constituent models of a supply chain in manufacturing, which exploits the inherent operating characteristics of its domain.

New method to minimize the preventive maintenance cost of series–parallel systems

General preventive maintenance model for input components of a system, which improves the reliability to ‘as good as new,’ was used to optimize the maintenance cost. The cost function of a maintenance policy was minimized under given availability constraint. An algorithm for first inspection vector of times was described and used on selected system example. A special ratio-criterion, based on the time dependent Birnbaum importance factor, was used to generate the ordered sequence of first inspection times. Basic system availability calculations of the paper were done by using simulation approach with parallel simulation algorithm for availability analysis. These calculations, based on direct Monte Carlo technique, were applied within the programming tool Matlab. A genetic algorithm optimization technique was used and briefly described to create the Matlab's algorithm to solve the problem of finding the best maintenance policy with a given restriction. Adjacent problem, which we called ‘reliability assurance,’ was also theoretically solved, concerning the increase of the cost when asymptotic availability value conforms to a given availability constraint.

A parallel algorithm for molecular dynamics simulation of branched molecules

To get an insight into the effects of molecular architecture in the behaviour of thin lubricant films we have devised an algorithm for simulation of branched molecules. We have used this algorithm successfully to simulate branched isomers of C 30. However the algorithm is flexible enough to be used for the simulation of more complex branched molecules. The resulting algorithm can be used in molecular dynamics simulation of branched molecules and could be helpful in designing new materials at the molecular level.

A Multiple-Heaps Algorithm for Parallel Simulation of Collision Systems

We consider the parallel simulation of collision systems. It has wide application, such as in hard-sphere molecular dynamics simulation for gas dynamics and crystals, as well as in studying molecular collision dynamics of chemical reactions. With detailed analysis, proper data structures are designed so that the central computational task is formulated as a consecutive search for the minimum in the collision time space of O(N2) entries, with multiple updates on O(N) entries in the same space per collision step. The abstraction and formulation enable us to incorporate efficient techniques in computer science into this application, which leads to a heap-based sequential algorithm of O(N log N) time in one typical collision step, where N is the number of particles of the simulated collision system. A parallel algorithm of multiple heaps with a diagonal-oriented mapping is then proposed. We show that the parallel algorithm is load balanced and the parallel time per collision step is O((N/P) log (N2/P)+log P), where P is the number of processors. The parallel algorithm uses two levels of partitioning independently, one in the particle-based physical space and the other in the collision time space. An exchange-shift communication algorithm is presented to bridge the two different partitioning schemes. Besides collision system simulation, the parallel multiple heaps algorithm may find applications in many other computing areas where a heap-based priority queue needs to be maintained, such as in fast level-set methods.

Analysis and Simulation of Mixed-Technology VLSI Systems

Circuit simulation has proven to be one of the most important computer aided design (CAD) methods for verification and analysis of, integrated circuit designs. A popular approach to modeling circuits for simulation purposes is to use a hardware description language such as VHDL. VHDL has had a tremendous impact in fostering and accelerating CAD systems development in the digital arena. Similar efforts have also been carried out in the analog domain which has resulted in tools such as SPICE. However, with the growing trend of hardware designs that contain both analog and digital components, comprehensive design environments that seamlessly integrate analog and digital circuitry are needed. Simulation of digital or analog circuits is, however, exacerbated by high-resource (CPU and memory) demands that increase when analog and digital models are integrated in a mixed-mode (analog and digital) simulation. A cost-effective solution to this problem is the application of parallel discrete-event simulation (PDES) algorithms on a distributed memory platform such as a cluster of workstations. In this paper, we detail our efforts in architecting an analysis and simulation environment for mixed-technology VLSI systems. In addition, we describe the design issues faced in the application of PDES algorithms to mixed-technology VLSI system simulation.

A Mixed Heuristic for Circuit Partitioning

As general-purpose parallel computers are increasingly being used to speed up different VLSI applications, the development of parallel algorithms for circuit testing, logic minimization and simulation, HDL-based synthesis, etc. is currently a field of increasing research activity. This paper describes a circuit partitioning algorithm which mixes Simulated Annealing (SA) and Tabu Search (TS) heuristics. The goal of such an algorithm is to obtain a balanced distribution of the target circuit among the processors of the multicomputer allowing a parallel CAD application for Test Pattern Generation to provide good efficiency. The results obtained indicate that the proposed algorithm outperforms both a pure Simulated Annealing and a Tabu Search. Moreover, the usefulness of the algorithm in providing a balanced workload distribution is demonstrated by the efficiency results obtained by a topological partitioning parallel test-pattern generator in which the proposed algorithm has been included. An extented algorithm that works with general graphs to compare our approach with other state of the art algorithms has been also included.

Parallel Simulations of Swirling Turbulent Flames

The feasibility of using massively paralleled computations as an engineering design tool is evaluated. A parallel Large-Eddy Simulation (LES) algorithm which simulates turbulent reacting flows using a space and time-accurate method, is used to model the complex flow found inside a realistic gas-turbine combustor. The parallelization philosophy and its implementation as a platform-independent solver is discussed. A performance analysis is carried out to determine the communication and storage requirements, and the associated overhead. As a case study, the LES methodology is used for a parametric investigation of swirl effects on the turbulent reacting flow in the gas-turbine.

A Parallel Algorithm for Simulation of Unsteady Flows in Multistage Turbomachinery

The numerical simulation of unsteady flow in multi-stage turbomachinery is computationally expensive. A parallel code using message-passing interface libraries, which was developed to reduce the turnaround time and the cost of computation, is presented in this paper. The paper describes the details of the parallel algorithm, including the parallelization rationale, domain decomposition and processor allocation, and communication patterns. The numerical algorithm was used to simulate the unsteady flow in a six-row test turbine. The numerical results present the salient features of the turbine flow, such as the temporal and spatial vari1 2 ation of velocity, pressure, temperature and blade force. To illustrate the advantages of parallel processing, the turnaround time of the turbine flow simulation was compared on several computers where the code was run in parallel and sequential mode. Nomenclature Cp Pressure coefficient F Force f Frequency p Pressure r Radius T Temperature γ Ratio of specific heats of a gas μ Viscosity ρ Density τ Skin friction ω Angular velocity Subscripts F Half-amplitude Fourier transform hub Hub location n− d Non-dimensional tip Tip location −∞ Upstream infinity Superscripts ∗ Total (or stagnation)

A parallel algorithm for simulation of large neural networks

The simulation of biologically realistic neural networks requires the numerical solution of very large systems of differential equations. Variables within the system can be changing at rates that vary by orders of magnitude, not only at different times of the solution, but at the same time in different parts of the network. Therefore, an efficient implementation must be able to vary the solution step size, and do so independently in different subsystems. A single processor algorithm is presented in which each neuron can be solved with its own step size by using a priority queue to integrate them in the correct order. But this leaves the problem of how communication and synchronisation between neurons should be managed when executing in parallel. The proposed solution uses an algorithm based on waveform relaxation, which allows groups of neurons on different processors to be solved independently and hence in parallel, for substantial parts of the computation. Realistic test problems were run on a distributed memory parallel computer and results show that speedups of 10 using 16 processors are achievable, and indicate that further speedups may be possible.

A multilevel unsymmetric matrix ordering algorithm for parallel process simulation

Computer simulation of complex chemical processes is increasingly being used in the design optimization and control of chemical facilities. Industrial-scale modeling involves the solution of large systems of algebraic differential equations. This is very computationally intensive with a large part of the computing time attributed to the repeated solution of large, sparse. unsymmetric systems of linear equations. One way of speeding up the simulation is to solve the linear systems efficiently in parallel by reordering the unsymmetric matrices into a bordered block-diagonal (BBD) form. In this paper a multilevel ordering algorithm is presented. A multilevel technique, which provides a global view, is combined with a Kernighan–Lin algorithm to form an effective unsymmetric matrix ordering algorithm MONET (Matrix Ordering for minimal NET-cut). Numerical results confirm that this algorithm gives ordering of better quality than existing algorithms.

Derivative estimates parallel simulation algorithm based on performance potentials theory

An efficient derivative estimates parallel simulation algorithm is presented based on the new Performance Potentials theory (Cao and Chen, 1997). Two main ideas are introduced: First, a new processor-partitioning pattern, Screwy Partitioning, which can make complete load balance on a time-costing simulation part; Second, modified Common Random Number, which can remove the large amount of broadcasting cost of sample path data at the price of only adding a very little workload. The simulation experiments on an SPMD parallel computer show that this algorithm can achieve near linear speedup.

Parallel molecular dynamics simulation: Implementation of PVM for a lipid membrane

This paper describes a parallel algorithm for Molecular Dynamics simulation of a lipid membrane using the isothermal—isobaric ensemble. A message-passing paradigm is adopted for interprocessor communications using PVM3 (Parallel Virtual Machine). A data decomposition technique is employed for the parallelization of the calculation of intermolecular forces. The algorithm has been tested both on distributed memory architecture (DEC Alpha 500 workstation clusters) and shared memory architecture (SGI Powerchallenge with 20 R10000 processors) for a dipalmitoylphosphatidylcholine (DPPC) lipid bilayer consisting of 32 DPPC molecules and 928 water molecules. For each architecture, we measure the execution time with average work load, and the optimal number of processors for the current simulation. Some dynamical quantities are presented for a 2 ns simulation obtained with 5 processors on DEC Alpha 500 workstations. Our results show that the code is extremely efficient on 5–8 processors, and a useful addition to other major computational resources.

An evaluation of implementations of the CMB parallel simulation algorithm on distributed memory multicomputers

A model of a message passing network is used to analyze the behavior of three implementations of the Chandy-Misra-Bryant (CMB) parallel simulation algorithm. The characteristics of the model, the organization of the logical processes (LPs) that constitute the simulator and the characteristics of the host parallel computer have a definite influence on the achieved performance, measured in terms of speedup. Large, loaded models help CMB to synchronize with a minimum overhead, efficiently exploiting the available parallelism. Mapping several LPs onto each processor achieves a better use of the available processing power, because while an LP is blocked (synchronizing) others can use the CPU. However, it is not convenient to map too many LPs onto each processor because the synchronization cost would be too high. The communication demands of CMB reduce its efficiency in environments where the cost of passing messages is too high: the performance of CMB running on a network of workstations is quite poor; in contrast, good speedups can be achieved using commercial multicomputers.

The application of scalable distributed memory computers to the finite element modeling of electromagnetic scattering

Large-scale parallel computation can be an enabling resource in many areas of engineering and science if the parallel simulation algorithm attains an appreciable fraction of the machine peak performance, and if undue cost in porting the code or in developing the code for the parallel machine is not incurred. The issue of code parallelization is especially significant when considering unstructured mesh simulations. The unstructured mesh models considered in this paper result from a finite element simulation of electromagnetic fields scattered from geometrically complex objects (either penetrable or impenetrable.) The unstructured mesh must be distributed among the processors, as must the resultant sparse system of linear equations. Since a distributed memory architecture does not allow direct access to the irregularly distributed unstructured mesh and sparse matrix data, partitioning algorithms not needed in the sequential software have traditionally been used to efficiently spread the data among the processors. This paper presents a new method for simulating electromagnetic fields scattered from complex objects; namely, an unstructured finite element code that does not use traditional mesh partitioning algorithms. © 1998 This paper was produced under the auspices of the U.S. Government and it is therfore not subject to copyright in the U.S.

Provably Good Partitioning and Load Balancing Algorithms for Parallel Adaptive N-Body Simulation

We present an efficient and provably good partitioning and load balancing algorithm for parallel adaptive N-body simulation. The main ingredient of our method is a novel geometric characterization of a class of communication graphs that can be used to support hierarchical N-body methods such as the fast multipole method (FMM) and the Barnes--Hut method (BH). We show that communication graphs of these methods have a good partition that can be found efficiently sequentially and in parallel. In particular, we show that an N-body communication graph (either for BH or for FMM) can be partitioned into two subgraphs with equal computation load by removing only $O(\sqrt{n\log n})$ and O(n2/3 (log n)1/3) number of nodes, respectively, for two and three dimensions. These bounds on node-partition imply bounds on edge-partition of $O(\sqrt{n}(\log n)^{3/2})$ and O(n2/3 (log n)4/3), respectively, for two and three dimensions. To the best of our knowledge, this is the first theoretical result on the quality of partitioning N-body communication graphs for nonuniformly distributed particles. Our results imply that parallel adaptive N-body simulation can be made as scalable as computation on regular grids and as efficient as parallel N-body simulation on uniformly distributed particles.

Modeling and Parallel Simulation of Large ATM Switch Fabrics

Simulation is an essential tool in the development, testing, and performance evaluation of communication systems. In this paper we discuss a practical case where large ATM switches are needed and develop a cell-level simulator to evaluate the performance of such switches. We exploit the conservative time-window algorithm for parallel simulation to reduce the simulation time and resolve the memory limitation problem with sequential machines. We simulate ATM switches of sizes up to 1 k x 1 k ports. A speed-up of 7 using ten processors on a shared-memory multiprocessor is achieved.

Parallel Processing And Fingerprint Image Comparison

A neural-network-based fingerprint image comparison (FIC) system is presented, and a variety of approaches for parallel simulation of the FIC system are explored. The FIC system contains an image preprocessing stage and a neural network decision stage. The preliminary training and testing results of the FIC system using the NIST fingerprint database lire discussed. The parallel simulation algorithms of the neural network presented in this paper include data parallel programming (SIMD) and message-passing programming (MIMD) implemented on the CM-5. The results of the performance analysis show the advantages of the MIMD approach: it is effective, flexible, and can avoid detailed data distribution mechanisms. On the other hand, the SIMD implementation presents worse performance than sequential because of the existence of extensive data routing overhead.

A parallel spectral Fourier-Nonlinear Galerkin algorithm for simulation of turbulence

We present a high-order parallel algorithm, which requires only the minimum interprocessor communication dictated by the physical nature of the problem at hand. The parallelization is achieved by domain decomposition. The discretization in space is performed using the Local Fourier Basis method. The continuity conditions on the interfaces are enforced by adding homogeneous solutions. Such solutions often have fast decay properties, which can be utilized to minimize interprocessor communication. In effect, the predominant part of the computation is performed independently in the subdomains (processors) or using only local communication. A novel element of the present parallel algorithm is the incorporation of a Nonlinear Galerkin strategy to accelerate the computation and stabilize the time integration process. The basic idea of this approach consists of decomposition of the variables into large scale and small scale components with different treatment of these large and small scales. The combination of the Multidomain Fourier techniques with the Nonlinear Galerkin (NLG) algorithm is applied here to solve incompressible Navier–Stokes equations. Results are presented on direct numerical simulation of two-dimensional homogeneous turbulence using the NLG method. © 1997 John Wiley & Sons, Inc. Numer Methods Partial Differential Eq 13: 699–715, 1997