Engineering Commons^™

Efficient Image Processing In Resource-Constrained Visual Sensor Networks, Hongtao Du

Doctoral Dissertations

Visual sensor networks (VSNs) that employ content-rich 2-D images or image sequences as the basic media have been evolving rapidly in recent years. Besides the critical resource constraints that are already inherent in any micro-sensor networks, the development of VSNs also faces challenges from device design, image transmission, and onboard image processing, among which efficient onboard processing is the most difficult to tackle. The focus of this dis- sertation is to develop efficient image processing solutions from three aspects: to improve the time-consuming image processing algorithms using pipelined and parallel computing; to dis- tribute the computation more effectively through novel …

Scheduling Algorithms For Scalable High-Performance Packet Switching Architectures, Xike Li

Doctoral Dissertations

Packet switching fabrics constitute a fundamental building block of all Internet routers. As a core technology, the switching engine is responsible for enabling multiple input (ingress) ports to be dynamically linked to output (egress) ports, thereby allowing packets to effectively traverse the router. Scheduling algorithms, which play a key role in switching fabrics, determine the dynamic configurations of the input-output matchings. The ever growing need for additional bandwidth and more sophisticated service provisioning in next- generation networks necessitates the introduction of scalable packet scheduling solutions that go beyond legacy schemes.

Switch architectures can be coarsely classified into two categories, in …

Accelerating Exact Stochastic Simulation Of Biochemical Systems, James Michael Mccollum

Doctoral Dissertations

The ability to accurately and efficiently simulate computer models of biochemical systems is of growing importance to the molecular biology and pharmaceutical research communities. Exact stochastic simulation is a popular approach for simulating such systems because it properly represents genetic noise and it accurately represents systems with small populations of chemical species. Unfortunately, the computational demands of exact stochastic simulation often limit its applicability. To enable next-generation whole-cell and multi-cell stochastic modeling, advanced tools and techniques must be developed to increase simulation efficiency. This work assesses the applicability of a variety of hardware and software acceleration approaches for exact stochastic …