Online video gaming has become a ubiquitous aspect of modern-day video gaming. It has gained immense popularity due to its accessibility and immersive experience, resulting in millions of players worldwide participating in various online games. Depending on the type of gameplay, the players’ quality of experience (QoE) in online video gaming can be significantly affected by network factors such as high bandwidth and low latency. As such, providers of online gaming services are competing to offer the highest quality of experience to their users at reasonable prices. To achieve this objective, online game providers face two main challenges. Firstly, they must accurately estimate the network throughput capacity required to meet the servers’ demands and ensure that the QoE is not compromised. Secondly, they must be able to secure the required throughput with network providers, which, in the current conventional network infrastructure, is neither agile nor dynamic. Thus, online game providers have to prepay for extra network throughput capacity or choose a cost-effective capacity that may result in potential QoE losses during peak usage. To address these challenges, this thesis proposes a deep neural network-based model that utilizes a QoE-aware loss function for predicting the future network throughput de- mand. The model can accurately estimate the network throughput capacity required to maintain QoE levels while minimizing the cost of network resources. By doing so, on- line game providers can achieve optimal network resource allocation and effectively meet servers’ demands. Furthermore, this thesis proposes a slice optimizer module that employs 5G network slicing and a machine learning model to optimize network slices in a cost-efficient manner that satisfies both the online game provider’s and the network provider’s requirements. This module can dynamically allocate network resources based on the game provider’s QoE requirements, the network provider’s resource availability, and the cost of network resources. As a result, online game providers can efficiently manage network resources, optimize network slicing, and effectively control the cost of network resources.

To understand remotely sensed data, one must understand the relationship between radiative transfer models and their predictions of the interaction of solar radiation on geophysical media. If it can be established that these models are indeed accurate, some form of evaluation has to be performed on these models, for users to choose the model that suits their requirements. This thesis focuses on the implementation of a variable linear kernel model, its validation, and to study its application in the prediction of BRDF effects using two different neural networks-- the backpropogation and the radial basis function neural network and finally to draw conclusions on which neural network is best suited for this model. Based on these results the optimum number of kernels for this model is derived.

A unified optimization framework for traffic signal priority scheduling based on the phasetime network models is presented in this research work with two mathematical programming formulations: (i) Mixed Integer Linear Programming (MILP) formulation and resilient MILP formulation (R-MILP). A heuristic algorithm is developed using the delay generated by the cumulative arrival and departure curves for searching optimum solution in phase-time network. An acceptance and rejection policy is also developed based on the proposed R-MILP. A set of numerical experiment with the proposed policy is conducted for fully adapted and coordinated phase-time network. The third set of numerical experiment is destined for the comparison of the performance of proposed phase-time network with the signal timing given by traditional traffic engineering method and Multi-Modal Intelligent Traffic Signal System (MMITSS) (1). The results show that in each case, the proposed formulation gives shorter delay and arterial travel time than the other two methods.

No description available.

No description available.

No description available.

No description available.

Ferrous objects in earth's magnetic field cause distortion in the surrounding ambient field. This distortion is a function of the object's material properties and geometry, and is known as the magnetic signature. As a precursor to first principle modeling of the phenomenon and a proof of concept, the goal of this research is to predict offboard magnetic signatures from on-board sensor data using a neural network. This allows magnetic signature analysis in applications where direct field measurements are inaccessible. Simulated magnetic environments are generated using MATLAB's Partial Differential Equation toolbox for a 2D geometry, specifically for a rectangular shell. The resulting data sets are used to train and validate the neural network, which is configured in two layers with ten neurons. Sensor data from within the shell is used as network inputs, and the off-board field values are used as targets. The neural network is trained using the Levenberg-Marquardt algorithm and the back propagation method by comparing the estimated off-board magnetic field intensity to the true value. This research also investigates sensitivity, scalability, and implementation issues of the neural network for signature estimation in a practical environment. / Electrical and Computer Engineering

The emerging passive optical network (PON) requires the burst mode clock and data recovery (BM-CDR) for the successful data detection, with a strict requirement in the locking time. Two innovative BM-CDR schemes are proposed, modeled, simulated, and analyzed. They simplify the circuit design and reduce the chip size and the power consumption by utilizing the characteristics of the optical components in the upstream fiber link. One scheme utilizes the phenomenon of the clock tone generation by the fiber dispersion. The other scheme utilizes the nonlinear relaxation oscillation of the directly modulated laser (DML) to generate the clock tone. The phenomenon of the clock tone generation by the DML relaxation oscillation is discovered for the first time. Both schemes do not incur extra cost, additional optical components or electrical circuit blocks. In both schemes, the BM clock recovery (CR) circuitry is based on the injection locked oscillator (ILO). Its behavior in the BM-CR application with the input of the distorted non-return-to-zero (NRZ) data is simulated at the system level for the first time. The BM-CR circuitry is designed and fabricated in a standard 0.18 !lm CMOS technology to experimentally demonstrate the two schemes operating at the bit rate close to 10 Gbps. / The emerging passive optical network (PON) requires the burst mode clock and data recovery (BM-CDR) for the successful data detection, with a strict requirement in the locking time. Two innovative BM-CDR schemes are proposed, modeled, simulated, and analyzed. They simplify the circuit design and reduce the chip size and the power consumption by utilizing the characteristics of the optical components in the upstream fiber link. One scheme utilizes the phenomenon of the clock tone generation by the fiber dispersion. The other scheme utilizes the nonlinear relaxation oscillation of the directly modulated laser (DML) to generate the clock tone. The phenomenon of the clock tone generation by the DML relaxation oscillation is discovered for the first time. Both schemes do not incur extra cost, additional optical components or electrical circuit blocks. In both schemes, the BM clock recovery (CR) circuitry is based on the injection locked oscillator (ILO). Its behavior in the BM-CR application with the input of the distorted non-return-to-zero (NRZ) data is simulated at the system level for the first time. The BM-CR circuitry is designed and fabricated in a standard 0.18 !lm CMOS technology to experimentally demonstrate the two schemes operating at the bit rate close to 10 Gbps. / Thesis / Doctor of Philosophy (PhD)

One of the major classes of polymer matrix resins under consideration for structural composite applications in the infrastructure and construction industries is the dimethacrylate matrix resin. An investigation of the relationships between the chemical structures and properties of these dimethacrylate/styrene networks has been conducted. Oligomer number average molecular weights of the polyhydroxyether ranging from 700 to 1200g/mole were studied with systematically varied styrene concentrations to assess the effects of crosslink density and chemical composition on glass transition temperatures, toughness, tensile properties and matrix strain. Network densities have been estimated from measurements of the rubbery moduli at Tg + 40°C. Within this rather small range in vinyl ester molecular weight, toughness of the resultant networks improved tremendously as the vinyl ester oligomer Mn was increased from 700g/mole to 1200g/mole due to improvements in the resistance to crack propagation. As styrene concentration was increased along all series' of materials, brittleness increased even though the molecular weight between crosslinks increased. This was attributed to the inherent relative brittleness of the polystyrene chemical structure relative to the polyhydroxyether component. This may also be related to the reactivity ratios dictating styrene and vinyl ester sequence length and warrants further investigation. As expected, the volume contraction upon cure also decreased significantly as styrene was decreased, and thus residual cure stresses may be reduced in fiber-reinforced composites. Vickers microhardness values decreased for each of the series when molecular weight increased and styrene content decreased. Two different cure procedures were compared to assess the effects of conversion on the physical and mechanical properties. All mechanical properties investigated (i.e. fracture toughness, tensile strength, and microhardness) were dependent on the cure procedure. Materials cured at 140°C were harder, more brittle, had lower elongations and higher rubbery moduli than those cured at 25°C followed by a 93°C postcure. A maximum in the degree of conversion occurred with increasing polymerization temperature and can be explained by the competition between the chemical reaction and molecular mobility. The overall shrinkage per moles of vinyl groups converted was the same when the materials were cured at 25°C or 140°C. However, in the room temperature cured samples, there was essentially no further densification of the network during postcure, regardless of the postcure temperature. A mono-methacrylate analogue of the dimethacrylate terminated poly(hydroxyether) oligomer was synthesized and copolymerized with styrene to study the effects of chain transfer during elevated temperature reactions. / Ph. D.