Showing papers on "Parallel processing (DSP implementation) published in 1978"

A Survey of Parallel Algorithms in Numerical Linear Algebra.

Abstract: The existence of parallel and pipeline computers has inspired a new approach to algorithmic analysis. Classical numerical methods are generally unable to exploit multiple processors and powerful vector-oriented hardware. Efficient parallel algorithms can be created by reformulating familiar algorithms or by discovering new ones, and the results are often surprising. A comprehensive survey of parallel techniques for problems in linear algebra is given. Specific topics include: relevant computer models and their consequences for programs, evaluation of arithmetic expressions, solution of general and special linear systems of equations, and computation of eigenvalues.

Parallel Processing of Power System Analysis Problems Via Simple Parallel Microcomputer Structures

Abstract: With the ever-increasing complexity of power systems, transient analyses are demanding a tremendous amount of computer effort. Recent advances in microprocessor technology have developed inexpensive processing units and memory as a set of small chips. This paper investigates the suitability of utilizing simple, extensible parallel microcomputer structures for efficient simulation of power system transients. Algorithms exploiting the sparsity structure of the equations involved are discussed and simulated results are presented.

Sorting, categorization, and visual search

Abstract: Publisher Summary It is probable that no two signals ever received by a human sense organ are ever precisely identical. This truism has nontrivial consequences for the understanding of human information processing. It implies that signals from the same source will have considerable variation and that this variation may indeed be easily discriminable. Early discussions of signal categorization and of visual search centered around discussions as to whether perceptual processing could best be regarded as a serial or a parallel process. In the event, both characteristics were claimed for perceptual processing in different tasks and at different stages of practice in each task. There have been very illuminating discussions of the extent to which data from particular experiments can be expected to provide evidence for the operation of systems of one of these classes or of the many subclasses. The distinctions between serial and parallel processing must entail distinctions between the types of classification a subject is required to make, or can make, when categorizing input.

An Analysis of Computational Cost in Image Processing: A Case Study

Abstract: Picture processing includes a variety of algorithms employed at different stages within a recognition process. Although single parallel algorithms have been suggested, no systematic comparison with sequential algorithms is available. We propose here some parallel algorithms solving a selected set of picture-processing tasks and compare their speed, when implemented on a CLIP machine, with that of the equivalent sequential algorithms implemented on a standard minicomputer. It is shown that parallel algorithms are faster than the sequential ones by a factor depending on the size of the digitization matrix, even when arithmetical operations are widely involved.

Incoherent optical ambiguity function generator

Abstract: An electro-optical system for performing the computation of the radar ambiguity function or other similar mathematical function using the parallel processing capability of optical systems including a photo-responsive, real-time programmable mask for storing a set of sample values {sp }. A light source is positioned to illuminate the mask and have its intensity modulated as a function of the sample values {sp }. The image from the programmable mask is transferred to a CCD photo-sensor array that is clocked in synchronism with the light pulse outputs of the light emitting diode to give rise to electro-optical products which are integrated in the CCD structure to yield the ambiguity or other similar mathematical function.

Dynamic graphics using quasi parallelism

Abstract: Dynamic computer graphics is best represented as several processes operating in parallel. Full parallel processing, however, entails much complex mechanism making it difficult to write simple, intuitive programs for generating computer animation. What is presented in this paper is a simple means of attaining the appearance of parallelism and the ability to program the graphics in a conceptually parallel fashion without the complexity of a more general parallel mechanism. Each entity on the display screen can be independently programmed to move, turn, change size, color or shape and to interact with other entities.The scheme presented herein begins with the notion of a quantum of time, or tick, within which there are no ordering constraints on events. Each entity or actor decides what it must do upon the next tick. Ticks are a powerful means of controlling parallel processes but are usually at too low a conceptual level for user convenience. Higher-level operations built upon the tick mechanism are presented, most notably the ability to instruct any entity or group of entities to gradually change or move at a rate that is itself changeable by the same operation. To illustrate these ideas a simple celestial mechanics simulation is presented. Upon each tick the velocities and positions of the objects are updated by the gravitational and propulsive forces acting upon them.Ticks are only one product of an object-oriented programming style. For the best control and the most modularity, graphics programming should be object oriented. Each object displayed, and its parts, should be independently programable. Instead of being passive data, objects should be responsible for the changes in their position or appearance. Instead of a global controller, each object should interact with the others.

A New Symbol Timing Recovery Technique for Burst Modem Applications

Abstract: A fast symbol timing recovery technique for PSK signals is presented. For the carrier recovery and for the symbol timing recovery subsystems, a parallel processing of the received modulated signal is employed. Theoretically it is shown that the optimum delay is 2/f_{c} and not T/2 as published in the literature, where f c and T are the carrier frequency and the symbol interval, respectively. This technique has the advantage over the currently employed serial processing technique in faster acquisition of symbol timing and with it in faster overall system synchronization. Experimental verification of theoretical results is presented and applications in satellite and terrestrial microwave communication systems are discussed.

Application of the residue number system to computer processing of digital signals

Abstract: The residue number system offers parallel processing, digital hardware, implementations for the binary operations of addition, subtraction and multiplication. This paper discusses the use of the residue number system in implementing digital signal processing functions, in which these binary operations abound. The paper covers implementations using arrays of read only memories, and briefly discusses the use of parallel microprocessor structures. ROM array implementations of scaling operations are also presented.

Parallel and Nondeterministic Time Complexity Classes (Preliminary Report)

Abstract: A random access machine model that has parallel processing and string manipulation is introduced. It is shown that NP is equal to the class of sets accepted by this model in nondeterministic time 0(log n), and this result is generalized to arbitrary time classes. A well-behaved subclass of the class of parallel, nondeter-ministic programs is analyzed.

The STARAN architecture and its application to image processing and pattern recognition algorithms.

Abstract: The ST ARAN E is a general purpose parallel computer. However, certain aspects of the STARAN E's architecture, specifically, a single instruction stream-multiple data stream organization, high speed I/O, a flip (permutation) network and a conditional operation capability on each parallel processing element, are particularly pertinent to the areas of image processing and pattern recognition. Some of the characteristics of image processing and pattern recognition which provide a good fit to the ST ARAN' s architecture are repetitiveness, spatial dependencies, complex parallel decision. making and high 110 to computation ratios. These characteristics and their corresponding STARAN E architectural accommodations are described in detail. The other aspects of ST ARAN' s architecture are described in only enough detail to provide a basis for discussion. *

Guest Editorial: Spatial Light Modulators

Abstract: The real-time and parallel processing advantages of coherent optical processors are well known. The ease with which the Fourier transform and correlation operations (that are the hallmark of data processing and pattern recognition) can be performed optically has intrigued researchers for years. This once virgin discipline of optical computing is now maturing and emerging from the laboratory and research arena into deliverable and fielded systems. As this occurs, one of the key technology items that continues to plague program managers and contractors is the real-time and reusable spatial light modulator (SLM) used as the input (and often filter plane) transducer. Without such devices, the full speed advantages of optical processors cannot be realized.

Corrections and Comments on "On Models for Diagnosable Systems and Probabilistic Fault Diagnosis"

Abstract: Dr. H. Fujiwara of Osaka University, Osaka, Japan, has brought to our attention two errors in the above paper.1

On some parallel programming techniques

Abstract: Some recent controversy over tightly coupled vs. loosely coupled parallelism may be due to not applying synergism effectively, Herein, parallelism (the execution of a plurality of applications in a plurality of processors) is analyzed from the applications point of view through software, in order to demonstrate that at least five distinct modes of parallelism can be synergistically combined, and some architectures proposed in the literature are surveyed to show embodiments of these modes. This point-of-view may be better in some design studies than that developed from the control structure or from the communication structure of a parallel system, or it may augment those points of view.

Searching for multiple targets simultaneously: is it really possible?

Abstract: In an attempt to clarify the findings of Neisser and his associates with respect to parallel processing of multiple target items in a visual search task, subjects were tested in a multiple-target search situation in which the various target sets were not nested. 20 sessions were run for each subject with 12 trials per day. The targets used were geometric forms, allowing for manipulation of the number of features in each form. The results do not confirm those of Neisser, in fact, the interaction of session × target-set size was in a direction opposite to that predicted by a parallel scan model. These findings indicate that what Neisser reported as parallel preattentive processing may in fact have reflected a tendency for subjects to scan for a master set of targets, regardless of which subset of that set was actually present on a given set.

Recognition Memory: Low cost content‐addressable parallel processor for speech data manipulation

Abstract: Associative memories for computers have heretofore been too costly, compared with conventional random access memories, for most applications Recognition Memory (REM) employs a judicious blend of serial and parallel processing to make available at low cost most of the advantages not only of associative memory but also, more broadly, of what C Foster calls Content Addressable Parallel Processors (Van Nostrand Reinhold, New York, 1976) REM accesses stored data by the content rather than the storage location addresses—though REM can be used in RAM mode with addressed read and write There are six hardware recognition functions: memory locations can be made to respond if their contents, with respect to the presented comparand, match exactly (=), or are ≠, >, ⩾, <, or ⩽ These are combinable by Boolean operators in software REM also has a multiwrite function whereby one instruction writes data into multiple locations—all responders to a prior recognize operation, say—and this, in combination with bit maskin

Parallel initial costate search algorithms for computing optimal controls

Abstract: This paper describes a collection of parallel initial costate search algorithms suitable for implementation on an advanced computer with the facility for large scale parallel processing. Specifically, a parallel nongradient algorithm and a parallel variable metric algorithm have been used to search for the initial costate vector that defines the solution to the optimal control problem. To avoid the computational problems sometimes associated with simultaneous forward integration of both the state and costate equations, a parallel shooting procedure based upon partitioning of the integration interval was considered. Application of this parallel procedure to a forced Van der Pol system indicates that convergence time is significantly less than that required by highly efficient serial procedures.

Pattern generating unit

Abstract: PURPOSE:To make processing with a high speed by using less quantity of hardware, through the simultaneous parallel processing at each memory cycle.

An algorithm for parallel processing of the state increment dynamic programming problem

Abstract: An algorithm is proposed for the parallel processing of the dynamic programming problem. The procedure combines the state increment approach with function partitioning and evaluate the partitioned cost function using the successive approximations technique as n one-dimensional problems in a decoupled form.

Waveform screening method of semiconductor elements

Abstract: PURPOSE:To screen acceptance and rejection by optically detecting the characteristic waveforms of the elements to be examined, storing picture element information and parallel processing the information with processing and judgement circuits prepared for the same number as that of the kinds of waveform shapes.

Information transfer system

Abstract: PURPOSE:To enable the information processing system enabling the simultaneous parallel processing for the ON-line processing by utilizing the same input and output bus line.

Communication issues in parallel algorithms and systems

Abstract: As multiple computer systems become more widely accepted, the importance of parallel programming increases. In this paper, approaches to the design and analysis of parallel algorithms are investigated. Through several examples the importance of interprocessor communication in parallel processing is demonstrated. Various techniques that are applicable in the design and analysis of parallel algorithms are examined with emphasis on those techniques that incorporate communication aspects. The paper discusses several models of synchronous and asynchronous parallel computation and their use in analyzing algorithms. Relatively primitive methodologies for designing parallel algorithms are discussed and the need for more general and practical methodologies is indicated.