I. Mitrani, Simulation Techniques for Discrete Event Systems, Cambridge University Press, 1982.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....estimation algorithm to determine when it is safe to garbage collect old history items, the estimated GVT value is used to commit irreversible le I O operations. In some senses, performing a VHDL simulation using the TyVIS and warped simulation kernels is similar to performing a process oriented [32] simulation. Process oriented views are useful for modeling but are dicult to implement eciently in a simulation system [35] The simulation of a collection of VHDL processes (post elaboration) can be viewed as implementing a process oriented simulation. A key issue in this type of simulation is ....

Isreal Mitrani. Simulation Techniques for Discrete Event Systems. Cambridge University Press, New York, 1982.

.... can be expressed as a linear combination p(x) fi 1 p 1 (x) fi 2 p 2 (x) fi n pn (x) P fi i = 1 ; fi i 0 ; 6) where p 1 (x) pn (x) are probability density functions, the following two step procedure can be applied to produce a sample from the total distribution p(x) (Mitrani, 1982): 1. Generate a random integer, m, being 1 with probability fi 1 , 2 with probability fi 2 and so on. 2. Generate a random variable from the probability density function pm (x) and let it be the output. The method for random number generation described above lends itself to a branch and bound ....

Mitrani, I. (1982). Simulation techniques for discrete event systems. Cambridge University Press.

....event, continuous simulation is appropriate. In this kind of simulation, the model is often a set of di#erential equations solved with numerical methods where time is one of the free variables. Continuous simulation generally causes problems with speed, numerical accuracy, and statistical accuracy [14]. These problems come from the algorithms used in the computation of random numbers as well as the algorithms used for solving numerical integrals. 4 This is the purpose of Sim94 A concurrent simulator for plan driven troops. 13] 4 SIMULATIONS 19 Simulation Time independent Time dependent ....

I. Mitrani. Simulation techniques for discrete event systems. Cambridge Computer Science Texts 14, 1982.

....was evaluated to false. 3.1 Simulation techniques One main purpose of a simulation program is to generate operation paths for the system being studied. This means an operation path is defined as an instance of a behaviour pattern and the system state as a function of time, as described in [Mit82]. We can separate between two possible simulation structures . fixed time increment: Also called synchronous simulation; the simulator clock is very similar to a real one and the value of the clock is incremented by a given (constant) amount. variable time increment: Also called asynchronous ....

....is the paradigm of eventdriven behaviour that can be simulated by a dynamic event list so that time skips from one event epoch to the next. Our approach provides asynchronous simulation using a clock with variable time increments. There are two ways to handle these clocks, as described in [BFL83, Mit82]. event orientation: For each possible event there is a procedure associated with it, which is invoked every time an event of this type occurs. process orientation: A process is defined as a sequence of events, together with a set of actions accompanying each event. An event sublist is ....

I. Mitrani: "Simulation techniques for discrete event systems", Cambridge Computer Science Texts, No. 14, Cambridge University Press, 1982.

....7 2. Introduction This manual is not intended as a tutorial on the concepts of simulation in general, but rather how to write simulations in the C SIM system. However, in order to be able to do this certain key simulation concepts will be briefly described. The interested reader is referred to [6] for detailed descriptions of these concepts and for further discussions on simulation modelling. 2.1. Simulation Models To model a system is to replace it by something which is: simpler and or easier to study. equivalent to the original in all important respects. Therefore, before ....

....pseudo random numbers, we continue to call the random number generators. The starting point for generating arbitrary distribution functions is to produce a standard uniform distribution. As we shall see, all other distributions can be produced based upon this. Interested readers are referred to [6] for a more complete treatment of this topic) All of the distribution functions in C SIM rely upon inheritance to specialise the behaviour obtained from the uniform distribution class. 5.1. RandomStream For historical reasons, the actual uniform distribution class is called RandomStream. This ....

I. Mitrani, "Simulation Techniques for Discrete Event Systems", Cambridge University Press, Cambridge, 1982.

....A survey on recent results of their quantitative analysis is found in the paper of Baccelli et al. 6] in these proceedings. Simulation is an important method for solving queueing models. In this paper, we will not discuss that approach. The interested reader is referred to the books of Mitrani [125], Sauer and MacNair [146] and Rubinstein [145] The paper is organized as follows. Section 2 contains a global discussion of some solution methods that have been successful in the performance analysis of parallel and distributed systems. Six key models for this performance analysis are discussed ....

I. Mitrani. Simulation Techniques for Discrete Event Systems. Cambridge University Press, Cambridge, 1982.

....with a highlevel modelling technique called transition classes. State spaces may be very large, Markov chains may be stiff, i.e. rare events may occur, simulations can be executed highly in parallel, and hybrid evaluation is possible. INTRODUCTION Monte Carlo simulation for discrete event models [Mit82] is a very popular and general technique in model based performance or reliability evaluation which is easily to apply. However, this simulation has some drawbacks, and problems appear. Some of them are addressed in this paper. Estimators may have a bias, and their variance may be high which leads ....

....simulation has some drawbacks, and problems appear. Some of them are addressed in this paper. Estimators may have a bias, and their variance may be high which leads to large confidence intervals. The last occurs when rare events are present. Ingenious variance reduction methods have been invented [HH64, Fis78, Mit82, Rip87, Ros89, McG92], e.g. conditioning, control variates, discrete time methods [HIS96] hybrid methods [SS84] antithetic variates, importance sampling, the restart technique [VAVA91] injection simulation [vM93] to enumerate a few. Some of them are difficult to apply, some are restricted to certain models, some ....

I. Mitrani. Simulation Techniques for Discrete Event Systems. Cambridge University Press, Cambridge, 1982.

.... can be expressed as a linear combination p(x) fi 1 p 1 (x) fi 2 p 2 (x) fi n pn (x) P fi i = 1 ; fi i 0 ; 5) where p 1 (x) pn (x) are probability density functions, the following two step procedure can be applied to produce a sample from the total distribution p(x) [5]: 1. Generate a random integer, l, being 1 with probability fi 1 , 2 with probability fi 2 and so on. 2. Generate a random variable from the probability density function p l (x) and let it be the output. Using this method recursively from the top node and downwards through the tree results in a ....

I. Mitrani. Simulation techniques for discrete event systems. Cambridge University Press, 1982.

....many modalities into one system. Common themes, such resource allocation and contention [22] are also missing from the works of the graphics community. Algorithms for resource contention are useful to manage scenes with many competing objects. Also absent are methods of analysis and validation [14]. These methods are important to verify that the outputs of animated scenes are correct and valid. 2.2.2 Computer Simulation Computer simulation seeks to create and to analyze programs that simulate the behaviors of real world systems. The field offers a strong vocabulary of terms, methods, and ....

I. Mitrani. Simulation Techniques for Discrete Event Systems. The Cambridge University Press, Cambridge, England, 1982.

....system, confirms the low overheads of this approach, and demonstrates its capability to simulate over one million processes in a process oriented model. 1 INTRODUCTION Three widely recognized world views for simulation are: event oriented view, process oriented view and activity scanning view (Mitrani 1982). The first two views are the more widely used among the three views. All the three views are equivalent in the sense that process oriented and activity scanning views can be translated into semantically equivalent event oriented views. An advantage of using the process oriented view is that ....

Mitrani I. 1982. Simulation Techniques for Discrete Event Systems. Cambridge University Press.

....time of 10,000,000 cycles. 18 5. 1 Simulation Experiment Design As is usual in simulation experiments, we choose to make a single long run, divide it into k portions (after having discarded the initial transient part) and then work with the sample of observations obtained from those portions [36]. Furthermore, high accuracy in the simulation results is required since the performances of some of the schemes to be compared are very close (they sometimes differ in the third decimal place or less) The simulation run is repeated K times with independent random number streams. Let X ji be the ....

I. Mitrani, "Simulation Techniques for Discrete Event Systems", Cambridge University Press, 1982, Cambridge, New York, chapter 5, pp.86-105.

....of complex systems and phenomena. The system or phenomenon being studied is modeled by a set of state variables and timestamped events. Execution of an event may change the state of the system and or schedule future events. There are several different paradigms for discrete event simulation [18]; we follow the process oriented, message based paradigm in this paper (discussed in more detail in section 2) A simulation protocol refers to a set of rules that are used to correctly execute a simulation. Discrete event simulations have traditionally been carried out using a sequential global ....

I. Mitrani. Simulation Techniques for Discrete Event Systems. Cambridge University Press, 1982.

....collect the data. Later on, if neural network approach were adopted in real industry, the data collection process should be done in real communication network operating circumstances. 3. 4 The Simulation Model Since we will use MMBP model, we consider that a discrete time simulation will fit best [20, 21]. The time axis is divided into equal intervals called slots. As cells arrive, they are stored in a fixed size buffer. Transmission (service) of a cell is synchronized to start only at slot boundaries. Cells which arrive in a slot are eligible for transmission at the beginning of the next slot. If ....

I. Mitrani, Simulation Techniques for Discrete Event Systems, Cambridge University Press, 1982.

....distribution of 1 ms. Simulations of the composed queuing network have been performed using the QNAP2 (Queuing Network Analysis Package) tool [9] The results have been obtained using standard techniques for analyzing simulation results, such as the approach of independent replications, see, e.g. [6]. Fig. 7 shows the sensitivity of the mean execution time of CC entities on the mean response time for handover requests for both scenarios. The BC time was fixed to 3 ms. The 95 confidence intervals for these simulations are given in Table 1; for the sensitivity of the BC execution time we refer ....

I. Mitrani. Simulation Techniques for Discrete Event Systems. Cambirdge University Press, 1982.

....Figure 4. Bursts of cell arrival to input. 4 Simulation Setup To cope with the high cost of long simulation runs, we choose to make a single long run, divide it into k portions (after having discarded the transient part) and then work with the sample of observations obtained from those portions [21]. The simulation run is repeated K times with independent random number streams. Let X ji be the ith observation in the jth run. For example, in a jth run, we obtained the batched means, X j1 , X j2 , X jk . Then the estimates of the mean and its variance are given by Mean = M( 1 ( ....

I. Mitrani, "Simulation Techniques for Discrete Event Systems", Cambridge University Press 1982, Cambridge, New York, Sec.5, pp. 86-105.

....the ratio of total implosion losses to NC Theta DP . The desired value for I is 0, i.e. no losses due to implosion. For a given set of parameters, we run between ten and twenty simulation runs, and the graphs shown here have a percentage offset of under Sigma4.9 with a confidence level of 95 ([14]) The multi thread support of the Simula language ( 15] was used to implement the threads mentioned in the section 3. We simulated the implosion losses in the following manner: we define a buffer at the parent for storing the incoming responses. An incoming packet is stored in the buffer if ....

I. Mitrani, Simulation Techniques for Discrete Event Systems, Cambrisge Computer Science Texts 14. Cambridge University Press, 1982.

....and equivalence checking are available. However, the analysis again involves the solution of large Markov chains. Since process algebras are a relatively new paradigm, only few specialized analysis techniques are available, as compared to stochastic Petri nets. ffl Simulation techniques [20] do not impose any restriction on the model design and provide an obvious technique for model analysis. However, the computational effort for simulating large scale models is immense 1 , especially if the system ex 1 The simulation of 5 minutes real time in a moderately complex CCSN in [27] ....

I. Mitrani. Simulation Techniques for Discrete Event Systems. Cambridge University Press, 1982.

....in the list, and Empty returns TRUE is the list is empty, FALSE otherwise. Clear removes all Link objects from the list. 3. Example Having considered the simulation package, we shall now show how it can be used by looking at an example. 3.1. Job Service Simulation This example is taken from [Mitrani 82] and simulates a process scheduler for a machine which attempts to execute as many process (jobs) as possible. The machine can only process one job at a time and job requests are queued until the machine can deal with them. However, the machine is prone to failures, and so jobs started will be ....

I. Mitrani, "Simulation Techniques for Discrete Event Systems", Cambridge University Press, Cambridge, 1982, p. 22.

No context found.

I. Mitrani, Simulation Techniques for Discrete Event Systems, Cambridge University Press, 1982.

No context found.

I. Mitrani. Simulation Techniques for Discrete Event Systems. Cambridge University Press, 1982.

No context found.

I. Mitrani, Simulation Techniques for Discrete Event Systems (Cambridge University Press, 1982).

No context found.

I. Mitrani. Simulation techniques for discrete event systems. Cambridge Computer Science Texts, 1982.

No context found.

I. Mitrani. Simulation techniques for discrete event systems. Cambridge University Press, 1982.

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