Discrete-event simulation appears to be an ideal candidate for parallel processing not only because many large-scale simulations take extremely long execution times on conventional computers but also because the systems being modelled often contain considerable amounts of intrinsic parallelism. Ability to simulate large models in a reasonable time is the motivation for seeking speed advantages offered by parallel computer systems. The Connection Machine is an example of a massively parallel computer with a general communications network in which any processor can communicate with any other that is well suited for the DEVS (Discrete Event System Specification) broadcast models simulation. However, a new approach is required to mapping the DEVS abstract simulator onto a SIMD architecture such as that of the Connection Machine CM-2. This dissertation extends the DEVS formalism to allow the exploitation of data parallelism afforded by a massively parallel SIMD architecture. A broadcast model simulation environment is implemented in *Lisp on the Connection Machine CM-2. Two examples of parallel processor models are presented for demonstration. Several runs are made on the implementation using the Parallel Processor Broadcast Architecture model. The experimental results are compared to those from DEVS-Scheme running on sequential machines such as the Sun-4 workstation and the Motorola Delta 88K MultiPersonal Computer using the same model. The results show that simulations on the CM-2 are approximately 150 times faster. In addition, to get some insight on the effect of increasing the number of processors of a SIMD architecture (i.e., the size of the broadcast model simulated), runs were made to measure the execution times. The execution results confirm our analysis pointing to imitations in the SIMD architecture for exploiting DEVS internal event parallelism.

We investigate the number of periodic points of certain discrete quadratic maps modulo prime numbers. We do so by first exploring previously known results for two particular quadratic maps, after which we explain why the methods used in these two cases are hard to adapt to a more general case. We then perform experiments and find striking patterns in the behaviour of these general cases which suggest that, apart from the two special cases, the number of periodic points of all quadratic maps of this type behave the same. Finally we formulate a conjecture to this effect.

As input ows of secondary raw materials show high volatility and tend to behave in a chaotic way, the identification of the main drivers of the dynamic behavior of returns plays a crucial role. Based on a stylized productionrecycling system consisting of a set of nonlinear difference equations, we explicitly derive parameter constellations where the system will or will not converge to its equilibrium. Using a constant elasticity of substitution production function, the model is then extended to enable coverage of real world situations. Using waste paper as a reference raw material, we empirically estimate the parameters of the system. By using these regression results, we are able to show that the equilibrium solution is a Lyapunov unstable saddle point. This implies that the system is sensitive on initial conditions that will hence impede the predictability of product returns. Small variations of production input proportions could however stabilize the whole system. (authors' abstract)

The report detailing the Hardware Accelerator for the JPEG encoder is organized into three sections. First, it will review the processes of the Joint Photographic Experts Group (JPEG) encoding and decoding standard. Second, it will review three different implementations of the discrete cosine transform in hardware. This is a very computationally intensive element of the JPEG encoding process and the analysis of these designs covers the benefits and costs of the various approaches for the Field Programmable Gate Array (FPGA) and Application Specific Integrated Circuit (ASIC) implementations. Finally, it will discuss this specific hardware accelerator design for a color state transformation for the standard JPEG encoder. An eight by eight matrix of Red, Green, Blue (RGB) values is passed into the FPGA as well as calculated in software. The Y Cr Cb results from that of the hardware accelerator implementation are compared with the software implementation for computational accuracy and the differences in computation time are sampled for a comparison. There is a clear 38% improvement in speed from the hardware accelerator. / text

A new method has been developed to incorporate particles with internal structure into the framework of the Variance Reduction method [17] for solving the discrete velocity Boltzmann Equation. Internal structure in the present context refers to physical phenomena like rotation and vibration of molecules consisting of two or more atoms. A gas in equilibrium has all modes of internal energy at the same temperature as the translational temperature. If the gas is in a non-equilibrium state, translational temperature and internal temperatures tend to proceed towards an equilibrium state during equilibration, but they all do so at different relaxation rates. In this thesis, rotational energy of a distribution of molecules is modeled as a single value at a point in a discrete velocity space; this represents the average rotational energy of molecules at that specific velocity. Inelastic collisions are the sole mechanism of translational and rotational energy exchange, and are governed by a modified Landau-Teller equation. The method is tested for heat bath simulations, or homogeneous relaxations, and one dimensional shock problems. Homogeneous relaxations demonstrate that the rotational and translational temperatures equilibrate to the correct final temperature, which can be predicted by conservation of energy. Moreover, the rates of relaxation agree with the direct simulation Monte Carlo (DSMC) method with internal energy for the same input parameters. Using a fourth order method for convecting mass along with its corresponding internal energy, a one dimensional Mach 1.71 normal shock is simulated. Once the translational and rotational temperatures equilibrate downstream, the temperature, density and velocity, predicted by the Rankine-Hugoniot conditions, are obtained to within an error of 0.5%. The result is compared to a normal shock with the same upstream flow properties generated by the DSMC method. Internal vibrational energy and a method to use Larsen Borgnakke statistical sampling for inelastic collisions is formulated in this text and prepared in the code, but remains to be tested. / text

No description available.

No description available.

This dissertation consists of three essays in labor economics. The first essay models how migrants crossing the border between the United States and Mexico respond to increases in border enforcement. We model a potential migrants' joint decision of whether to cross the border and, if so, where to cross the border using a random utility function. Our model allows us to calculate the migrants' substitution patterns: does more enforcement primarily on one part of the border primarily deter migrants from crossing the border altogether, or simply divert them to other parts of the border? We find that a substantial proportion of migrants are indeed diverted. These findings should serve as a caveat to policy makers who seek to address immigration reform issues primarily through tightening the border.The second chapter models the internal migration decisions of U.S. households during the period 935 to 1940. We measure the impact of spending on New Deal programs on migration patterns. Using a model of random utility similar to that in prior chapter, we find that more public works and relief spending in a region made it more attractive to potential migrants, while additional spending on the Agricultural Adjustment Administration (AAA) made the locale less attractive. The structural nature of our model allows us to compute counterfactual estimates to assess the overall impact of these programs. We find that regional disparities in spending on public works and relief programs we responsible for nearly 20% of long distance moves made between regions during this period.In the third chapter, we decompose the gap between mean sentences for males and females in the U.S. criminal justice system into the portion that can be explained by differences in the average severity of the crime committed by males and females and the portion explained by differences in how males and females who commit the same crime are treated. We find that differences in characteristics of the defendant can explain only half of the gap between mean male and females sentences, suggesting that women receive more lenient treatment in the U.S. criminal justice system.

The body of literature investigating spectator attendance in sports has developed using various forms of regression with secondary data, leading to a series of fairly consistent findings. The aggregate secondary data that has been used in these studies, which has been effective in developing this body of knowledge, is inherently limited in its ability to explain the most basic element of the attendance issue: how individual consumers choose whether or not to attend a game. The objective of this research is to provide these consumer-level insights, specifically for games in the Canadian Football League, by utilizing a discrete choice methodology. This study generates primary data from actual consumers, while incorporating many of the standard demand determinants. The results of this study show that while the quality of both participating teams are significant predictors of demand, their relative quality is not; contradicting the uncertainty of outcomes hypothesis.

Appointment scheduling involves picking a strategy for sequencing patient appointments such that the wait time per patient is minimized and the idle time and overtime for the doctor are minimized. The goal of this project is to increase the number of scans per day at the echocardiogram clinic in the Hospital for Sick Children. The objectives were realized by performing simulations of the workflow of the echo clinic using simulation software. The simulation model did not precisely reflect the echo clinic, and the disparity was attributed to limitations in the simulation model. Nevertheless, the user accepted the model and six different policy change scenarios were explored. All six scenarios yielded significant increases in average scans per day per sonographer. Scenario IV, which eliminated the use of sonographer schedules, was recommended to be implemented due to ease of implementation and lack of negative effect on patient wait time and sonographer overtime.