BE Computer Engineering (IOE, TU) Simulation and Modeling (IOE, CT 751 / ENCT 353) Question Paper 2079

(a) Define a system and a model. With suitable examples, distinguish between physical and mathematical models, and between static and dynamic models. [6]

(b) "Simulation is the appropriate technique only when the system is too complex for analytical treatment." Discuss this statement, clearly stating the situations when simulation is the right tool and when it is not appropriate. [6]

Consider a single-server queuing system (M/M/1). Customers arrive with inter-arrival times of 3, 1, 4, 2, 5 minutes and require service times of 2, 4, 1, 3, 2 minutes respectively for the first five customers.

(a) Explain the event-scheduling approach to discrete-event simulation and define the terms event, event list, system state and simulation clock. [5]

(b) Construct the simulation table by hand for the five customers and compute the average waiting time in queue, the average time a customer spends in the system, and the server utilization. [9]

(a) Describe the linear congruential method for generating pseudo-random numbers. State the conditions (Hull-Dobell theorem) under which a multiplicative-additive LCG achieves the maximum period. [6]

(b) Using the inverse-transform technique, derive a procedure to generate a random variate from an exponential distribution with mean 1/lambda. Hence generate one exponential variate using the random number R = 0.25 and lambda = 0.5. [6]

(a) Differentiate between verification and validation of a simulation model. Describe at least three techniques used to verify a simulation program and three techniques used to validate a model against the real system. [8]

(b) Compare general-purpose programming languages with special-purpose simulation languages (such as GPSS) for building simulation models, listing two advantages and two disadvantages of each. [4]

Apply the Kolmogorov-Smirnov test to test the following sequence of numbers for uniformity at the 0.05 level of significance (critical value D(0.05) = 0.565 for n = 5): 0.44, 0.81, 0.14, 0.05, 0.93. State your conclusion.

For an M/M/1 queue with arrival rate lambda = 6 customers/hour and service rate mu = 8 customers/hour, derive and compute: (a) the server utilization rho, (b) the expected number of customers in the system L, and (c) the expected waiting time in the queue Wq.

Explain the acceptance-rejection technique for generating random variates. Using a neat algorithm, show how it can be used to generate a random variate from a distribution whose density is bounded over a finite interval.

Differentiate between continuous and discrete simulation, and between deterministic and stochastic simulation, giving one real-world example of each type.

What properties must a good sequence of pseudo-random numbers possess? Briefly describe the runs test (test for independence) and explain what hypothesis it is used to verify.

With a neat flowchart, explain the general structure (the time-advance mechanism and the main steps) of a next-event time-advance discrete-event simulation.

Define the following terms with reference to system modeling: (a) entity, (b) attribute, (c) activity, (d) state of a system, and (e) endogenous and exogenous events.

Write short notes on any TWO of the following:

(a) Monte Carlo simulation

(b) Features of a good simulation software package

(c) Calibration of a model in the validation process