Parallel Computational Structure Research Articles

THE PARALLEL-PIPELINED IMPLEMENTATION OF THE FRACTAL IMAGE COMPRESSION FOR RECONFIGURABLE COMPUTING SYSTEMS

The developed fractal image compression method, implemented for reconfigurable computing systems is described. The main idea parallel fractal image compression based on parallel execution pairwise comparison of domain and rank blocks. Achievement high performance occurs at the expense of simultaneously comparing maximum number of pairs. Implementation fractal image compression for reconfigurable computing systems has two critical resources, as number of input channels and FPGA Look-up Table (LUT). The main critical resource for fractal image compression is data channels, and implementation this task for reconfigurable computing systems requires parallel-pipeline computations organization replace parallel, preliminarily produced performance reduction parallel computational structure. The main critical resource for fractal image compression is data channels, and implementation this task for reconfigurable computing systems requires parallel-pipeline computations organization replace parallel computations organiation. For using parallel-pipeline computations organization, preliminarily have produce performance reduction parallel computational structure. Each operator has routed to computational structure sequentially (bit by bit) to save computational resources and reduces equipment downtime. Storing iterated functions system coefficients for image encoding has been introduced in data structure, which correlates between corresponding parameters the numbers of rank and domain blocks. Applying this approach for parallel-pipeline programs allows scaling computing structure to plurality programmable logic arrays (FPGAs). Task implementation on the reconfigurable computer system Tertius-2 containing eight FPGAs 15 000 times provides performed acceleration relatively with universal multi-core processor, and 18 – 25 times whit to existing solutions for FPGAs.

Синхронный алгоритм параллельной реконструкции трехмерных изображений на основе системы обмена сообщениями MPI

Одни из наиболее важных вычислительных задач разработка и адаптация итерационных методов для решения сверхбольших разреженных систем алгебраических уравнений. К таким вычислительным задачам приводит задача итерационной параллельной реконструкции трехмерных изображений промышленных изделий. Важно, что итерационные методы решения вычислительных задач большой размерности реализуются на параллельных структурах намного эффективнее, чем прямые методы их решения. В этой работе описан синхронный параллельный алгоритм, основанный на использовании системы MPI для решения задачи реконструкции трехмерных изображений промышленных изделий. Purpose. Currently, one of the most important tasks is the development and adaptation of iterative methods for solving ultra-large sparse systems of algebraic equations. Such computational problems are caused by the iterative parallel reconstruction of three-dimensional images of industrial products. It is important that iterative methods for solving computational problems of large size are implemented on parallel structures much more efficiently than direct methods for solving them. This paper describes a synchronous parallel algorithm based on the MPI system for solving the problem of reconstruction of three-dimensional images of industrial products. Methodology. It is important that iterative methods for solving computational problems of larger dimensionality on parallel structures are implemented more efficiently than the direct ones. Most algorithms based on direct solving methods have a significant hereditary sequential structure and require a large number of processors interactions which cannot be executed in parallel mode. Iterative methods, for the most part, require a significantly smaller number of interactions of this type and are relatively easily mapped onto parallel computational structures. Equally important, in most cases parallel implementations of classical iterative methods are more effective in terms of computational speed. The parallel execution of the algorithm is based on the distribution of the process in some way between various groups of processors. Depending on the interaction method between local processors, two different types of parallel iterative algorithms execution are distinguished: synchronous and asynchronous. In the former case, it is assumed that the processors complete the calculations and exchange all the necessary results before the start of a new iteration. The main disadvantage of synchronous parallel algorithms is that they require synchronization of iterations. This is a very difficult task, especially with large number of processors. In addition, the overall calculation speed is limited by the speed of the slowest processor. At the same time, faster processors spend most of their time in the waiting mode. But, nevertheless, the implementation of these parallel algorithms can be effectively achieved using the MPI standard. Findings. Synchronous parallel computing algorithms and program codes for threedimensional tomographic reconstruction in a conical beam were developed. Program code debugging and numerical calculations were performed on a hybrid cluster based on the OpenPOWER architecture using the MPI system. For designing a parallel threedimensional tomographic reconstruction, a voxel form of parallelism was used. Originality. Implemented parallel iterative technology of three-dimensional images reconstruction has undeniable advantages over traditional sequential iterative tomographic reconstruction. It allows reducing the time of tomographic reconstruction as many as tens of times, provides the ability to reconstruct products with sizes of 5123 to 10243 voxels with simultaneous storage of the submatrix node of the projection matrix in RAM, which eliminates the need for recalculation of matrix coefficients for each new iteration.

Parallel Computational Structure of Noisy Quantum Circuits Simulation

We present the detailed description of parallel computational structure of quantum circuits modeling. The deep theoretical and experimental analysis of corresponding algorithms and relations of their features to the nature of quantum computations are considered. Special attention is paid to the extension of modeling to the case of noisy circuits, which appear in realistic quantum computers.

Optimal and Suboptimal Resource Sharing Schemes for Underlaid D2D Communications

Being attracted by offloading local services to adjacent users, the operator allows two proximity users to build up a direct communication link, which is a new supplementary to current cellular network and called device-to-device (D2D) communication. Due to the low power radiation and occupation of the licensed band, D2D users can share the resource with cellular users fully controlled by cellular network to improve the local frequency efficiency. Due to the combinatorial and nonconvex nature of formulated problem, solving such problem optimally is a nearly impossible task. However, we propose an upper bound computation method, which can be seen as a benchmark for resource sharing scheme design. We also propose another two suboptimal schemes, i.e., suboptimal resource sharing scheme I with parallel computation structure which can achieve optimality in the most cases, and suboptimal resource sharing scheme II with relatively low complexity at the cost of acceptable performance loss. Moreover, through extensive numerical results, we find that both suboptimal resource sharing schemes can outperform the existed works. To conclude, our work actually provide a tradeoff between complexity and performance.

Efficient patch-based approach for compressive depth imaging.

We present efficient camera hardware and algorithms to capture images with extended depth of field. The camera moves its focal plane via a liquid lens and modulates the scene at different focal planes by shifting a fixed binary mask, with synchronization achieved by using the same triangular wave to control the focal plane and the pizeoelectronic translator that shifts the mask. Efficient algorithms are developed to reconstruct the all-in-focus image and the depth map from a single coded exposure, and various sparsity priors are investigated to enhance the reconstruction, including group sparsity, tree structure, and dictionary learning. The algorithms naturally admit a parallel computational structure due to the independent patch-level operations. Experimental results on both simulation and real datasets demonstrate the efficacy of the new hardware and the inversion algorithms.

IM6D

We propose IM6D, a novel real-time magnetic motion-tracking system using multiple identifiable, tiny, lightweight, wireless and occlusion-free markers. It provides reasonable accuracy and update rates and an appropriate working space for dexterous 3D interaction. Our system follows a novel electromagnetic induction principle to externally excite wireless LC coils and uses an externally located pickup coil array to track each of the LC coils with 5-DOF. We apply this principle to design a practical motion-tracking system using multiple markers with 6-DOF and to achieve reliable tracking with reasonable speed. We also solved the principle's inherent dead-angle problem. Based on this method, we simulated the configuration of parameters for designing a system with scalability for dexterous 3D motion. We implemented an actual system and applied a parallel computation structure to increase the tracking speed. We also built some examples to show how well our system works for actual situations.

A client/server-based parallel coevolutionary algorithm for parallel machines scheduling problem with penalties

A problem of n independent jobs with different ready times and due dates to be scheduled on m parallel machines which aims at minimizing the total tardiness penalty costs of all jobs is considered. The decomposition and combination characteristics of the problem are studied. Based on these two characteristics, a parallel coevolutionary algorithm (PCA) is proposed. The PCA is implemented in a client/server (C/S)-based parallel coevolutionary computation structure, in which the server is responsible for global evolution by task assignment and the client is responsible for local evolution by task ordering. According to this structure, a genetic algorithm and a virus evolutionary genetic algorithm are developed, which are executed on server and client, respectively. The simulated experiments are designed, the comparison results with adapted algorithms show that the proposed algorithm C/S-PCA has superior performances both in convergence speed and in searching optimality.

Design of silicon brains in the nano-CMOS era: Spiking neurons, learning synapses and neural architecture optimization

We present a design framework for neuromorphic architectures in the nano-CMOS era. Our approach to the design of spiking neurons and STDP learning circuits relies on parallel computational structures where neurons are abstracted as digital arithmetic logic units and communication processors. Using this approach, we have developed arrays of silicon neurons that scale to millions of neurons in a single state-of-the-art Field Programmable Gate Array (FPGA). We demonstrate the validity of the design methodology through the implementation of cortical development in a circuit of spiking neurons, STDP synapses, and neural architecture optimization.

Parallel PSO-Based Optimal Strategy Study of Energy Efficient Operation Control for Train

Different from existing evolutionary algorithms which usually are implemented in serial computation mode, two improved parallel particle swarm optimization algorithms based on parallel computation structure model are proposed to search the energy efficient operation strategy for a train in railway network. The algorithms are designed and verified using a simulation case. Compared with other methods, these improved parallel PSO-based algorithms have a much better performance in efficiency and may be considered to put into practice.

Parallel Implementation of OBDD-Based Splitting Surface Search for Power System

Parallel computational structure is helpful for many complicated problems, especially those which can be divided into multiple independent simpler sub-problems. The ordered binary decision diagrams (OBDD)-based splitting surface search algorithm owns this kind of dividability, derived from the associative law of Boolean expression and the dividability of matrix operation. We have implemented the algorithm with the parallel computation structure MPI to save up computing time. Customized verification is available on our interactive illustrative website http://obdd.cfins.au.tsinghua.edu.cn/ (login user name: mag, password: 199707).

An MRI‐dedicated parallel FDTD scheme

AbstractThis study extends our recent works on object‐oriented Finite‐Difference Time‐Domain (FDTD) simulations into a parallel‐computing framework, producing substantially improved performance with respect to computing time efficiency. The simulator is intended to be a complete, high‐performance FDTD model of an MRI system including all temporal radio frequency (RF) and low‐frequency field generating units and electrical models of the patient. The parallel computational structure has been designed using a Message‐Passing Interface (MPI) Library. The implementation of the MRI‐dedicated FDTD algorithms within the parallel computing framework is detailed. To improve the simulator performance in a realistic network‐computing environment, several software design aspects have been investigated. These include: data transmission with consideration of buffered and synchronous communication, load balancing on the computing network using a weighting mechanism. The power of the optimized FDTD parallel architecture is illustrated by two distinct, large‐scale field calculation problems involving the study of the interaction of RF‐fields with human tissue (whole‐body, male and female) in high field MRI and characterization of the temporal eddy currents induced on the cryostat vessel during gradient switching. The examples demonstrate the improved capabilities of the simulator. In addition, through the parallelization of the FDTD algorithm we are in a position to study various MRI phenomena at much higher spatial resolutions than was previously feasible. © 2007 Wiley Periodicals, Inc. Concepts Magn Reson Part B (Magn Reson Engineering) 31B: 147–161, 2007

Technique for synthesizing specialized parallel computer structures for digital image processing

This paper proposes a technique for synthesizing specialized parallel computational structures for digital image processing. The technique is based on the use of graph models and is oriented toward the use of programmable integrated logic circuits as a set of components.

Estimation and tracking of quasi-periodically varying processes

The problem of identification/tracking of quasi-periodically varying processes is considered. Two solutions to this problem are presented, First the global search algorithm is derived. Then its decoupled version. with a highly parallel computational structure, is proposed.

Time Cost Analysis of Parallel Structures with Multi-Communication Nodes in a Shared Memory Environment

In the paper the time costs of several parallel computation structures are analyzed. These analyses are based on the assumptions that the processes communicate implicitly via the shared memory and that a locking mechanism is imposed on the access to shared variables. In previous work, an approach to estimating a set of special parallel computation structures has been developed. In this paper, we expand this approach and propose a new technique for covering more general parallel computation structures.

Parallel computation structures for a class of cyclic spectral analysis algorithms

A digital frequency smoothing algorithm for cyclic spectral analysis is described. Partitioning schemes are presented that have properties amenable to parallel implementation. Parallel computation structures, based on the partitioning schemes, are described and system architectures proposed.

Petri net models of fuzzy neural networks

Artificial neural networks (ANN's) are highly parallel and distributed computational structures that can learn from experience and perform inferences. Petri nets, on the other hand, provide an effective modeling framework for distributed systems. The basic concepts of Petri net are utilized to develop ANN-like multilayered Petri net architectures of distributed intelligence having learning ability. A Petri net model of single neuron is presented. A two-layer Petri net model-neural Petri net (NPN)-that uses this neuron model as a building block is described. A new class of Petri nets called the fuzzy neural Petri net (FNPN) is defined. The FNPN can be used for representing a fuzzy knowledge base and for fuzzy reasoning. Some application examples for the two Petri net based models are given. >

Recursive high-resolution reconstruction of blurred multiframe images

An approach to obtaining high-resolution image reconstruction from low-resolution, blurred, and noisy multiple-input frames is presented. A recursive-least-squares approach with iterative regularization is developed in the discrete Fourier transform (DFT) domain. When the input frames are processed recursively, the reconstruction does not converge in general due to the measurement noise and ill-conditioned nature of the deblurring. Through the iterative update of the regularization function and the proper choice of the regularization parameter, good high-resolution reconstructions of low-resolution, blurred, and noisy input frames are obtained. The proposed algorithm minimizes the computational requirements and provides a parallel computation structure since the reconstruction is done independently for each DFT element. Computer simulations demonstrate the performance of the algorithm.

Parallel computation structure for magnetohydrodynamic problems

Parallel computation structure for magnetohydrodynamic problems

Feature extraction in future detectors

As part of CERN's LAA detector development project, we have studied the possibilities of using modern fine-grain parallel computer structures for extracting high-level information (physics features) from fast detectors, in real time. Such information could subsequently be substituted for raw data and thus alleviate bandwidth problems, or could serve for “intelligent” triggering. The main objective is to show how fast today's digital devices, implemented as custom-made or commercial processors, can execute basic algorithms. The possibilities of integrating them efficiently in the data gflow of experiments, and the extrapolations to be made for technological evolution over the coming years will be critical parameters when judging realistic applications in the future.

Parallel Computing Structures for UD Decomposition with a One-dimensional Processor Array

In this paper, two parallel computing structures with a linear array of N processor element's (PEs) are presented which are specifically devised to perform UD decomposition of N x N dense and sparse matrices. Each PE in the structure is connected to a local memory in which the contents of a column are stored. The operations of all the PEs are identical and simple. Our proposed structures for dense and sparse matrices can attain (N 2 +3N−4)/2 and Z+N−2 steps, respectively, where Z is the number of nonzero elements along and upper the diagonal including fill-ins. Moreover, a computationally effective architecture and a reforming algorithm for the architecture are also shown.