H.W. Lawson, Parallel Processing in Industrial Real-Time Applications. Prentice Hall, 1992.  Home/Search   Document Not in Database   Summary   Related Articles   Check

.... satisfying C(q 1 , q 2 , q n ) The foregoing requirements, namely, that the parameters of S be measured, new parameters be computed and applied, and a new (desired) state of equilibrium be reached, all within T time units, suggest that the computation needs to be performed in real time [4, 8, 13]. A fundamental property of real time computation is that certain operations must be performed by speci ed deadlines. Thus, one or more of the following conditions may be imposed: 1. Each received input (or set of inputs) must be processed within a certain time after its arrival. 2. Each output ....

H.W. Lawson, Parallel Processing in Industrial Real-Time Applications, Prentice Hall, Englewood Clis, New Jersey, 1992.

....another is the total length of time required to go from START to FINISH. An implicit deadline is imposed by the source which needs to be told, by the end of time interval i, what values to send at the beginning of time interval i 1. Practical problems satisfying these properties are described in [10, 18, 19, 20, 31]. 4 Discrete steepest descent We begin we a simple metaphor. A skier at the top of a mountain wants to reach the valley as quickly as possible. The fastest way to accomplish this is to take the path of steepest descent. All other routes lead to the same destination but take more time. Now back ....

H.W. Lawson, Parallel Processing in Industrial Real-Time Applications , Prentice Hall, Englewood Cliffs, New Jersey, 1992.

..... We also wish to emphasize that our definition, while striving to be as general as possible, is particularly suitable for our purposes in this paper. Many other more or less specialized definitions exist; see, for example, the various interpretations of the notion of real time provided in [6, 12, 18]. We assume in what follows that the inputs received in real time are deposited in a buffer from where they are picked up for processing. The size of this buffer is fixed and independent of the model of computation used. 3 SUPERLINEAR SPEEDUP Speeding up the sequential solutions to ....

H.W. Lawson, Parallel Processing in Industrial Real-Time Applications , Prentice Hall, Englewood Cliffs, New Jersey, 1992.

....the RS interface of the system is used. Generally, time triggered systems require an increased effort in the design phase of the system, but provide an easier verification of the temporal correctness [14] Since time triggered systems are designed according to the principle of resource adequacy [15], it is guaranteed that sufficient computing resources are available to handle the speci6 fied peak load scenario. On the other hand, timetriggered systems are often blamed for their bad flexibility. The STI overcomes this limitation by introducing means to configure the interaction of the ....

H. W. Lawson. Parallel Processing in Industrial Real-Time Applications. Prentice Hall, Englewood Cliffs, New Jersey, 1992.

....and a flat view to the nodes by a linear 64 bit address scheme (compare [2] The cookie table enumerates the nodes and stores the unique identifier and the base address of each device in the network. structures change dramatically if 0,1 or a limited number larger than 1 have to be managed [51]. It depends on the device class requirements if locking mechanisms or management functions for mutual exclusion have to be implemented in the driver. A different topic is whether the driver is able at its level of abstraction to decide if the access policy is used or not. Both cases can not be ....

....static resolution. In the PURE system, the approach of Aspect Oriented Programming [48] AOP) is used. Management structures change dramatically in complexity depending on the cardinality of elements. It has to be distinguished between 1, a fixed number N or a fully dynamic number of elements n [51]. The same principle holds for polymorphism if the number of derived classes drops to 1. The tool PUMA [57] analyses the selected PURE components for the cardinal number of the used components and changes the management structures accordingly. The TEReCS tool [11] is the configurator for the ....

Harold W. Lawson. Parallel Processing in Industrial Real-Time Applications. Prentice Hall, Englewood Cliffs, NJ, 1992.

....analytically or with the help of simulation. There is extensive experimentation going on in the design and configuration of both the processors and the interconnection network. For a general introduction to parallel computation one may consult, among others, the books by Hwang [51] and Lawson [60], on the design of hardware, and by Akl [3] and Bertsekas and Tsitsiklis [11] on the design of parallel scientific algorithms. 3 Although parallel programming is still far from being standardized, the basic desirable parallel computing modes have been characterized. One can attempt a broad ....

H.W. Lawson. Parallel processing in industrial real-time applications. Prentice--Hall Inc., 1992.

..... We also wish to emphasize that our definition, while striving to be as general as possible, is particularly suitable for our purposes in this paper. Many other more or less specialized definitions exist; see, for example, the various interpretations of the notion of real time provided in [6, 12, 18]. 3 SUPERLINEAR SPEEDUP Speeding up the sequential solutions to computational problems is the principal motivation behind parallel processing. In order to determine the goodness of a parallel algorithm that solves a certain problem, a measure known as speedup is used. Speedup is defined as the ....

H.W. Lawson, Parallel Processing in Industrial Real-Time Applications , Prentice Hall, Englewood Cliffs, New Jersey, 1992.

....several projects [1 2] treat modular robot concepts, focusing on modularity in actuators, joints and linkage. For control systems, various parallel processing approaches have been advocated, also considering the need for real time performance and increased dependability, for an overview see [3]. One interesting approach is the use of mechatronic modules in which sensor(s) and or actuator(s) are integrated with local processing units that incorporate a standardized communication interface. Such modules could then be interconnected mechanically and through the communication subsystem. ....

....performed in parallel. The algorithm is then said to be relaxed or partitioned, depending on whether no, or some, Paper to be presented at the International Symposium on Robotics and Manufacturing, August 1994 ISRAM 94. 5 interaction is needed between the processes to perform the algorithm, [3]. The trade offs and possibilities associated with partitioned relaxed algorithm decomposition have been treated in several disciplines, for instance, subsystem decoupling in large scale systems [9] and convergence of numerical algorithms in a distributed environment [10] A model for ....

H. Lawson, Parallel Processing in Industrial Real--Time Applications, Prentice Hall 1991.

....in decomposition. This facilitates design, and improves the system s modularity and maintainability, see further Section 3. Obviously LOC D approaches appear to be a very good starting point, that comes close to stated principles of resource adequacy (Ramseyer, et al. 1979; Duffie, 1982; Lawson, 1991), for example by implying that a time critical function should be provided within its own node. However, a pure LOC D approach is only possible with a suitable hardware structure. The hardware structure may require functions that are local with respect to one joint for example, to be implemented ....

....the controlled objects often constitute multivariable systems in which there may be dynamic couplings between several actuators (subsystems) Provided that such couplings are sufficiently weak, decomposition is straightforward. The resulting decomposed functions can then be said to be relaxed (Lawson, 1991). In other cases, the compensation for such couplings may require complex computations and information exchange to be performed, e.g. in robotics. Investigation of dynamic couplings is important because it is in some cases possible to simplify computations and reduce communications between ....

[Article contains additional citation context not shown here]

Lawson H. (1991). Parallel Processing in Industrial Real-Time Applications, Prentice Hall.

....23, 28, 38, 53, 54, 62, 66] It is also important to note that our definition, while striving to be as general as possible, is particularly suitable for our purposes in this paper. Many other definitions exist; see, for example, the various interpretations of the notion of real time provided in [10, 42, 65]. 2.2 Models of Computation Two models, one sequential and one parallel, are applied to the solution of the various real time computational problems studied in this paper. 2.2.1 Sequential Model This is the conventional model of computation used in the design and analysis of sequential (or ....

H.W. Lawson, Parallel Processing in Industrial Real-Time Applications , Prentice Hall, Englewood Cliffs, New Jersey, 1992.

....structure representing the input graph before the computation begins. This is not the case in a real time paradigm, where the data arrive as the algorithm proceeds. This is the basic idea underlying all real time computations, regardless of the various particular characteristics each of them has [3, 4, 5, 6, 7, 8, 16, 22]. In our formulation of the paradigm, a subset of the input data is given at the beginning of the computation, and the rest arrive at regular intervals. The algorithm must then deal with the newly arrived data and incorporate them into the solution computed so far, before a new chunk of data ....

H.W. Lawson, "Parallel Processing in Industrial Real-Time Applications", Prentice Hall, Englewood Cliffs, New Jersey, 1992.

....semaphores and monitors are examples of mutual exclusion mechanisms that guarantee the indivisible execution of a critical section and thereby the consistency of shared data. The following subsections give a short overview of these mechanisms, further details can be found in [Sil94] Bur97] [Law92] and [Tho98] CHAPTER 2 28 2.3.1 SEMAPHORES A semaphore S is an integer variable that, apart from initialization, is accessed only through two standard atomic operations: wait and signal. The semantics for these operations is as follows: Wait(S) If the value of the semaphore S is greater ....

Lawson, H., W., Parallel Processing in Industrial Real-Time Applications, Prentice-Hall International, U.S.A.

....systems exist. Lawson defines a realtime system to be . a system that assures that controlled activities progress and that stability is maintained and further, that the values of outputs and the time at which the outputs are produced are important to the proper functioning of a system. Law92] A more general definition is presented by Young, who defines a real time system to be: any information processing activity or system which has to respond to externally generated input stimuli with a finite and specified period. You82] 4 J. Hansson and M. Berndtsson Hence, a ....

Harold W. Lawson. Parallel Processing in Industrial RealTime Applications. Prentice-Hall, Inc., 1992.

....between the local memories of the modules. The processing and communication of data is done only via these local memories in the modules. Figure 2. System architecture To get a time deterministic system three different design rules are followed: 1) a resource adequate design strategy is used [Law92], 2) the programs, both for processing and communication, in the modules are executed cyclically [LS93] and (3) allocation of system resources is done in the design phase and not in the execution phase. Real time behaviour is achieved by tuning the program cycles in the modules so they fit the ....

Lawson H.W., with contributions by B. Svensson and L. Wanhammar. Parallel Processing in Industrial Real-Time Applications. Prentice-Hall, Englewood Cliffs, NJ, USA, 1992.

....22, 27, 37, 52, 53, 61, 65] It is also important to note that our definition, while striving to be as general as possible, is particularly suitable for our purposes in this paper. Many other definitions exist; see, for example, the various interpretations of the notion of real time provided in [9, 41, 64]. 2.2 Models of Computation Two models, one sequential and one parallel, are applied to the solution of the various real time computational problems studied in this paper. At the outset we state explicitly the following basic assumption: The analyses in this paper assume that all models of ....

H.W. Lawson, Parallel Processing in Industrial Real-Time Applications, Prentice Hall, Englewood Cliffs, New Jersey, 1992.

....in Paper C and the first realization of a REMAP module, a reconfigurable bit serial processor array with SIMD control, is described in Paper D. To allow real time operations and time determinism the architectural concept is based on resource adequacy, both in processing and communication [15, 16]. Learning algorithms are cyclically executed in distributed SIMD nodes, which access their data from local real time databases, updated with data from the other nodes via a shared high speed optical link. Other aspects of the project are described in [3, 4, 18, 19, 21, 23] The support for the ....

Lawson, H. W., Parallel Processing in Industrial Real-Time Applications, Englewood Cliffs, NJ, USA: Prentice-Hall, 1992.

No context found.

H.W. Lawson, Parallel Processing in Industrial Real-Time Applications. Prentice Hall, 1992.

....execution phases are not clearly separable and there is a need for an easy interconnect between development system and target system. To achieve time determinism, in order to guarantee realtime response, the approach taken in this paper is the principle of resource adequacy and cyclic execution [3], 4] Resource adequacy means that enough resources, both for computation and communication, are allocated statically in the design phase. However, as described above, the design and execution phases may alternate) Cyclic execution implies guaranteed timely response to stimuli. The program ....

Lawson H.W., with contributions by B. Svensson and L. Wanhammar. Parallel Processing in Industrial Real-Time Applications. Prentice-Hall, Englewood Cliffs, NJ, USA.

....risk analysis, reduction and minimization of all aspects are vital. As an important step towards risk minimization, a holistic view has been taken in the development of BASEMENT. The need for a holistic philosophy, as well as the impact of not having such a philosophy has been described by Lawson [23, 25]. In respect to the holistic view, risk reduction and minimization in computer based systems is accomplished by a combination of various means. Via robust architectures and design, via verifiable specifications, via thorough hazard analysis and hazard elimination, via the incorporation of safety ....

H.W. Lawson. Parallel Processing in Industrial Real-Time Applications. Prentice-Hall, 1992. ISBN 0-13-654518-1.

....risk analysis, reduction and minimisation of all aspects are vital. As a vital step towards risk minimisation, a holistic view has been taken in the development of BASEMENT. The need for a holistic philosophy, as well as the impact of not having such a philosophy has been described by Lawson [21, 23]. In respect to the holistic view, risk reduction and minimisation in computer based systems is accomplished by a combination of various means. Via robust architectures and design, via verifiable specifications, via thorough hazard analysis and hazard elimination, via the incorporation of safety ....

H.W. Lawson. Parallel Processing in Industrial Real-Time Applications. PrenticeHall, 1992. ISBN 0-13-654518-1.

No context found.

H.W. Lawson, Parallel Processing in Industrial Real-Time Applications , Prentice Hall, Englewood Cliffs, New Jersey, 1992.

No context found.

Harold W. Lawson, "Parallel Processing in Industrial Real-time Applications", Prentice-Hall, 1992.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC