Sensor networks are an important emerging class of networks that have many applications. A sink in these networks acts as a bridge between the sensor nodes and the end-user (which may be automated and/or part of the sink). Typically, convergecast is performed in which all the data collected by the sensors is relayed to the sink, which in turn presents the relevant information to the end-user. Interestingly, some applications require the sink to relay just a function of the data collected by the sensors. For instance, in a fire alarm system, the sinks needs to monitor the maximum of the temperature readings of all the sensors. For these applications, instead of performing convergecast, we can let the intermediate nodes process the data they receive, to significantly reduce the volume of traffic transmitted and increase the rate at which the data is collected and processed at the sink: this is known as in-network computation. Most of the current literature on this novel technique focuses on asymptotic results for large networks and for very elementary functions. In this dissertation, we study a new class of functions for which we want to compute explicit solutions for networks of practical size. We consider the applications where the sink is interested in the first M statistical moments of the data collected at a certain time. The k-th statistical moment is defined as the expectation of the k-th power of the data. The M=1 case represents the elementary functions like MAX, MIN, MEAN, etc. that are commonly considered in the literature. For this class of functions, we are interested in explicitly computing the maximum achievable throughput including routing, scheduling and queue management for any given network when in-network computation is allowed. Flow models have been routinely used to solve optimal joint routing and scheduling problems when there is no in-network computation and they are typically tractable for relatively large networks. However, deriving such models is not obvious when in-network computation is allowed. Considering a single rate wireless network and the physical model of interference, we develop a discrete-time model for the real-time network operation and perform two transformations to obtain a flow model that keeps the essence of in-network computation. This model gives an upper bound on the maximum achievable throughput. To show the tightness of that upper bound, we derive a numerical lower bound by computing a feasible solution to the discrete-time model. This lower bound turns out to be close to the upper bound proving that the flow model is an excellent approximation to the discrete-time model. We then adapt the flow model to a wired multi-rate network with asynchronous transmissions on links with different capacities. To compute the lower bound for wired networks, we propose a heuristic strategy involving the generation of multiple trees and effective queue management that achieves a throughput close to the one computed by the flow model. This cross validates the tightness of the upper bound and the goodness of our heuristic strategy. Finally, we provide several engineering insights on what in-network computation can achieve in both types of networks.

This thesis describes and evaluates how extending Intel's Concurrent Collections (CnC) programming model can address the problem of hybrid programming with high performance and low energy consumption, while retaining the ease of use of data-flow programming. The CnC model is a declarative, dynamic light-weight task based parallel programming model and is implicitly deterministic by enforcing the single assignment rule, properties which ensure that problems are modelled in an intuitive way. CnC offers a separation of concerns by allowing algorithms to be expressed as a two stage process: first by decomposing a problem into components and specifying how components interact with each other, and second by providing an implementation for each component. By facilitating the separation between a domain expert, who can provide an accurate problem specification at a high level, and a tuning expert, who can tune the individual components for better performance, we ensure that tuning and future development, such as replacement of a subcomponent with a more efficient algorithm, become straightforward. A recent trend in mainstream desktop systems is the use of graphics processor units (GPUs) to obtain order-of-magnitude performance improvements relative to general-purpose CPUs. In addition, the use of FPGAs has seen a significant increase for applications that can take advantage of such dedicated hardware. We see that computing is evolving from using many core CPUs to ``co-processing" on the CPU, GPU and FPGA, however hybrid programming models that support the interaction between multiple heterogeneous components are not widely accessible to mainstream programmers and domain experts who have a real need for such resources. We propose a C-based implementation of the CnC model for enabling parallelism across heterogeneous processor components in a flexible way, with high resource utilization and high programmability. We use the task-parallel HabaneroC language (HC) as the platform for implementing CnC-HabaneroC (CnC-HC), a language also used to implement the computation steps in CnC-HC, for interaction with GPU or FPGA steps and which offers the desired flexibility and extensibility of interacting with any other C based language. First, we extend the CnC model with tag functions and ranges to enable automatic code generation of high level operations for inter-task communication. This improves programmability and also makes the code more analysable, opening the door for future optimizations. Secondly, we introduce a way to specify steps that are data parallel and thus are fit to execute on the GPU, and the notion of task affinity, a tuning annotation in the specification language. Affinity is used by the runtime during scheduling and can be fine-tuned based on application needs to achieve better (faster, lower power, etc.) results. Thirdly, we introduce and develop a novel, data-driven runtime for the CnC model, using HabaneroC (HC) as a base language. In addition, we also create an implementation of the previous runtime approach and conduct a study to compare the performance. Next, we expand the HabaneroC dynamic work-stealing runtime to allow cross-device stealing based on task affinity. Cross-device dynamic work-stealing is used to achieve load balancing across heterogeneous platforms for improved performance. Finally, we implement and use a series of benchmarks for testing the model in different scenarios and show that our proposed approach can yield significant performance benefits and low power usage when using a hybrid execution.

The research study involved using the modified FEMWATER code to investigate the impact of total stress changes on groundwater flow in the vicinity of a salt tailings pile. Total stress and pore-pressure data observed at the Lanigan and Rocanville potash-mine sites were used to assist the development of a generic FEMWATER model. The original 3-D mesh considered for model study covers a region of 7.6 km x 7.6 km x 60 m. The simulated pile itself covers a surface area of 1.6 km x 1.6 km within the region. Symmetry of the idealized system allowed half of the system to be modelled to reduce the size of the mesh. The model was layered to facilitate different materials representing different hydrostratigraphic scenarios. The GMS-release of the FEMWATER code (version 2.1) was modified to simulate the pore-pressure response to total stress changes caused by tailings pile loading at the ground surface to be modelled. The modified code was verified before applying to present study.<p>Long-term pore pressure generation and dissipation due to pile construction was investigated for eleven hydrostratigraphic scenarios consisting of plastic clays, stiff till and dense sand layers commonly found in Saskatchewan potash mining regions. The model was run for two distinctive pile loading patterns. Model results indicated that the loading pattern has a significant influence on pore pressure generation beneath the pile. The model was initially run for 30 year pile construction period and later simulated for 15, 25 and 35 year construction periods to investigate the impact of loading rate. These results showed that, as expected, the peak pore water pressure head is proportional to the pile construction rate. A sensitivity analysis, which was carried out by changing hydraulic conductivity of stiff till, revealed that the lower the hydraulic conductivity, the greater the pore pressure generation beneath the pile.<p>Overall, the research study helped to understand and predict the influence of pile construction and hydrostratigraphy on pore-pressure changes beneath salt tailing piles. Low K/Ss or cv materials (compressible tills) demonstrate a slow dissipation rate and high excess pressures. Compared to dense sand which has very high K/Ss, till has very low K/Ss which causes in high excess pore pressure generation. Sand layers act as drains, rapidly dissipating pore pressures. Thicker low K/Ss units result in slower dissipation and higher pressures. As the thickness of the low K/Ss layer increases, the peak pressures increase as the drainage path lengthens. Thin plastic clay layers give rise to the highest pressures.<p>The model study showed that hydrostratigraphic scenarios similar to those found at Saskatchewan potash mine sites can generate the high pore pressures observed in the vicinity of salt tailings piles as a result of pile loading. Peak pressures are very sensitive to pile construction rates, loading patterns and hydrostratiagraphy of the region. Peak pressures can reach levels that would be of concern for pile stability on the presence of adverse geological conditions.

After semiconductor industry, TFT-LCD technology has become the other valuable technique for Taiwan to play an important role in the world. Both government and investors have already spent lots of effort to make it grow during last decade. To be able to increase the competitive, we are looking forward to merge the TFT-LCD manufactories. This paper focuses on the market expectation of merge of TFT-LCD industry in Taiwan. This paper has use two stock market public corporate, A.U.O. and Q.D.I, as our study and observation models and by applies the L-G model from Kermit D. Larson and Nicholas J. Gonedes, we will be able to prove the concept and examine our observation cases. On the other hand, I use the free cash flow model from Damodaran to evaluate the capital value for both companies. In particular, we examine them in three different scenarios to analysis their firm-value and use them as the ratio for future stock merge between both companies. By simulate our observation models and experiment for their fair market value; we will be able to demonstrate the L-G model is appropriate in this particular circumstance. According to the result and demonstration of cases study, it can provide the new prospective of Taiwan TFT-LCD industry in the future.

In order to achieve enhanced and cost-effective performance of engineering components, Surface Engineering embraces traditional and innovative surface technologies which modify the surface properties of metallic and non-metallic engineering components for specific and sometime unique engineering purposes. The surface roughness of an engineered surface may be classified as: the random surface roughness which is a product of surface finishing and the deterministic surface roughness which is engineered to increase the lubrication characteristics of the hydro dynamically lubricated thrust ring. The effect of stochastic/random roughness can not be ignored when the roughness is of the same amplitude as that of fluid film thickness. Average flow model derived in terms of flow factors which are functions of the roughness characteristics is used to study the random surface roughness effects on hydrodynamic lubrication of deterministic asperity. In addition, the effect of boundary conditions on flow factors is studied by calculating the pressure and shear flow factor using two different new boundary conditions. The results are obtained for random surface roughness having a Gaussian distribution of roughness heights.

Sensor networks are an important emerging class of networks that have many applications. A sink in these networks acts as a bridge between the sensor nodes and the end-user (which may be automated and/or part of the sink). Typically, convergecast is performed in which all the data collected by the sensors is relayed to the sink, which in turn presents the relevant information to the end-user. Interestingly, some applications require the sink to relay just a function of the data collected by the sensors. For instance, in a fire alarm system, the sinks needs to monitor the maximum of the temperature readings of all the sensors. For these applications, instead of performing convergecast, we can let the intermediate nodes process the data they receive, to significantly reduce the volume of traffic transmitted and increase the rate at which the data is collected and processed at the sink: this is known as in-network computation. Most of the current literature on this novel technique focuses on asymptotic results for large networks and for very elementary functions. In this dissertation, we study a new class of functions for which we want to compute explicit solutions for networks of practical size. We consider the applications where the sink is interested in the first M statistical moments of the data collected at a certain time. The k-th statistical moment is defined as the expectation of the k-th power of the data. The M=1 case represents the elementary functions like MAX, MIN, MEAN, etc. that are commonly considered in the literature. For this class of functions, we are interested in explicitly computing the maximum achievable throughput including routing, scheduling and queue management for any given network when in-network computation is allowed. Flow models have been routinely used to solve optimal joint routing and scheduling problems when there is no in-network computation and they are typically tractable for relatively large networks. However, deriving such models is not obvious when in-network computation is allowed. Considering a single rate wireless network and the physical model of interference, we develop a discrete-time model for the real-time network operation and perform two transformations to obtain a flow model that keeps the essence of in-network computation. This model gives an upper bound on the maximum achievable throughput. To show the tightness of that upper bound, we derive a numerical lower bound by computing a feasible solution to the discrete-time model. This lower bound turns out to be close to the upper bound proving that the flow model is an excellent approximation to the discrete-time model. We then adapt the flow model to a wired multi-rate network with asynchronous transmissions on links with different capacities. To compute the lower bound for wired networks, we propose a heuristic strategy involving the generation of multiple trees and effective queue management that achieves a throughput close to the one computed by the flow model. This cross validates the tightness of the upper bound and the goodness of our heuristic strategy. Finally, we provide several engineering insights on what in-network computation can achieve in both types of networks.

Bribie Island is a large, heterogeneous, sand barrier island that contains groundwater aquifers of commercial and environmental significance. Population growth has resulted in expanding residential developments and consequently increased demand for water. Caboolture Shire Council (CSC) has proposed to increase groundwater extraction by a new borefield. Two aquifers exist within the Quaternary sandmass which are separated by an indurated sand layer that is ubiquitous in the area. A shallow aquifer occurs in the surficial, clean sands and is perched on the indurated sands. Water levels in the shallow water table aquifer follow the topography and groundwater occurs under unconfined conditions in this system. A basal aquifer occurs beneath the indurated sands, which act as a semi-confining layer in the island system. The potentiometric surface of the basal aquifer occurs as a gentle groundwater mound. The shallow groundwater system supports water-dependent ecosystems including wetlands, native woodlands and commercial pine plantations. Excessive groundwater extraction could lower the water table in the shallow aquifer to below the root depth of vegetation on the island. Groundwater discharge along the coastline is essential to maintain the position of the saline water - fresh groundwater boundary in this island aquifer system. Any activity that changes the volume of fresh water discharge or lowers the water table or potentiometric surface below sea level will result in a consequent change in the saline water – freshwater interface and could lead to saline water intrusion. Groundwater level data was compared with the residual rainfall mass curve (RRMC) on hydrographs, which revealed that the major trends in groundwater levels are related to rainfall. Bribie Island has a sub-tropical climate, with a mean annual rainfall of around 1358mm/year (Bongaree station). Mean annual pan evaporation is around 1679mm/year and estimates of the potential evapotranspiration rates range from 1003 to 1293mm/year. Flows from creeks, the central swale and groundwater discharged from the area have the potential to affect water quality within the tidal estuary, Pumicestone Passage. Groundwater within the island aquifer system is fresh with electrical conductivity ranging from 61 to 1018ìS/cm while water near the coast, canals or tidal creeks is brackish to saline (1596 to 34800ìS/cm). Measurements of pH show that all groundwater is acidic to slightly acidic (3.3-6.6), the lower values are attributed to the breakdown of plant material into organic acids. Groundwater is dominated by Na-Cl type water, which is expected in a coastal island environment with Na-Cl rainfall. Some groundwater samples possess higher concentrations of calcium and bicarbonate ions, which could be due to chemical interactions with buried shell beds while water is infiltrating to depth and due to the longer residence times of groundwater in the basal aquifer. A steady-state, sub-regional groundwater flow model was developed using the Visual MODFLOW computer package. The 4 layer, flow model simulated the existing hydrogeological system and the dominant groundwater processes controlling groundwater flow. The numerical model was calibrated against existing data and returned reasonable estimates of groundwater levels and hydraulic parameters. The model illustrated that: .. The primary source of groundwater recharge is infiltration of rainfall for the upper, perched aquifer (Layer 1). Recharge for the lower sand layers is via vertical leakage from the upper, perched aquifer, through the indurated sands (Layers 2 and 3) to the semi-confined, basal aquifer (Layer 4). .. The dominant drainage processes on Bribie Island are evapotranspiration (15070m3/day) and groundwater seepage from the coast, canals and tidal creeks (9512m3/day). Analytical calculations using Darcy’s Law estimated that approximately 8000m3/day of groundwater discharges from central Bribie Island, approximately 16% less than the model. .. As groundwater flows preferentially toward the steepest hydraulic gradient, the main direction of horizontal groundwater flow is expected to be along an eastwest axis, towards either the central swale or the coastline. The central swale was found to act as a groundwater sink in the project area.

Master of Agribusiness / Department of Agricultural Economics / Allen M. Featherstone / Bananas are the number one selling produce item in the grocery store. On average, bananas account for 6% of produce department sales and 1% of total grocery store sales. According to The Packer’s “2010 Fresh Trends”, 88% of consumers in all categories purchase bananas. Also, 94% of consumers in the study purchased bananas within the last twelve months. Over the last decade, fuel prices have increased to a point where logistics and shipping have become more important than ever to the banana industry. This logistics challenge is compounded because there are no bananas grown in the United States and the fruit has to be shipped from around the world. Fuel is used at high rates via the ocean cargo and trucking shipments to meet yearly demand. To manage these logistical challenges, this thesis analyzes the optimal shipping route for bananas arriving to the west coast from Central and South America to various markets using a transshipment model. The goal of the transshipment model estimates the supply chain that creates the lowest cost. Through analysis of fuel, trucking, and shipping markets, the model makes the optimal decision regarding transportation routing. The model is limited to transportation costs only. However, items such as fruit costs and other additional up charges could be analyzed.

Contaminant fate and transport are two of the most important issues project officers have to deal with when developing a sound remediation strategy for a subsurface contamination site. To accurately assess these issues, knowledge of possible pathways, travel times and groundwater receptors are required. A groundwater flow model of a site facilitates the assessment process by determining flow paths, discharge areas and travel time from a contaminant source to a potential receptor. The resulting model can also be used to show potential impacts on drinking water sources and surface habitats. This project is Phase II of the proposed three phase project, Groundwater Flow Modeling of the Aquifer System at Ft. Belvoir (Widdowson, 1998). Phase I consisted of developing a conceptual model of the aquifer system, recommending a modeling strategy and developing a data collection strategy. The objectives of this phase are to design and construct a computer simulation of the groundwater flow system in the aquifers below Ft. Belvoir and to develop a strategy for improved data collection using the results of the model. Steps in this phase included creation of the numerical model, calibration to known water surface elevations, and a sensitivity analysis of the boundary conditions. The numerical model was created in the Department of Defense Groundwater Modeling System environment using MODFLOW. The model was calibrated to pre-1970 wells for the deep aquifer and recent site characterization wells for the upper aquifer. The head distribution was influenced the greatest by topography and the major creeks. Accuracy of the well surface elevations played a major role in the calibration process, as well as tidal influences. A sensitivity analysis showed that adjusting the recharge and the seepage face parameters affected the model results (head and groundwater flow rates) the greatest. While adjusting the constant head and general head boundaries affected the model results the least. / Master of Science

Computational and modelling skills are vital to most fields of biological research, yet traditional biology majors have no or little opportunity to develop these skills during their undergraduate education. We describe an approach, which can address this issue by a synergy of online resources called MathBench modules and Stock and Flow modelling. Using a step-by-step method starting with a MathBench ‘bootcamp’, we were able to achieve a significant gain in quantitative skills of students with no previous experience with model building. At the end of the course, the students were able to construct and analyse complex models and gained confidence in mathematical skills.