2012 Electrical and Chemical Engineering Research Opportunities
Prof. Wendi Heinzelman
Department of Electrical and Computer Engineering
wheinzel@ece.rochester.edu
Mobile Access to Cloud Services
The amount of data processed annually over the Internet has crossed the zetabyte boundary, yet this data cannot be efficiently processed or stored using today's mobile devices. However, advances in mobile access and cloud computing have brought the state-of-the-art in mobile-cloud computing to an inflection point, where the right architecture may allow mobile devices to run applications utilizing large amounts of data and processing. In this project, we will look at how to support mobile access to cloud services, focusing on how to optimally partition tasks between the mobile device, an intermediate computing device called a cloudlet, and the cloud servers. Using Android phones as the development platform, we will measure the latency and cost for different transmission medium such as Bluetooth, WiFi and cellular transmissions, as well as measure the latency and cost for accessing different cloud servers. Additionally, we will develop algorithms for determining what pre-processing can be done at the mobile or the cloudlet and how the processing can be accelerated through communication between cloud servers.
Protocol Development using Software Defined Radios
Mobile ad hoc networks provide the ability to connect large groups of people without the need for a pre-existing infrastructure. In such networks, each device is not only a source and a sink, but also a router. These networks are especially useful in military or disaster relief situations where networks need to be established quickly and efficiently and cannot rely on existing infrastructure. We have developed a suite of protocols called TRACE to support the real-time transmission of voice, video and data over a mobile ad hoc network. In this project, we will implement the TRACE protocols on software-defined radios called SORA. These radios will allow us to perform real-world tests on the different features of TRACE, including packet delivery and latency values, as well as providing an estimate of energy consumption, in different practical environments.
Prof. Thomas B. Jones
Department of Electrical and Computer Engineering, and Materials Science
jones@ece.rochester.edu
Research Project:
The goal of this project is to develop a wireless-based sensor system for on-line monitoring of electric power flow in buildings. The system will measure AC electrical current and voltage, log this data, and transmit it by a wireless network to a server facility for storage, on-demand display, and analysis. The specific objective of the project will be a system for real-time monitoring of AC currents, voltages, and time-average power in one of the 3-phase mains in the Hopeman Engineering Building. For the hardware, we will rely largely upon off-the-shelf components, including current transformers, power meters, and an Arduino board enhanced by wireless and A/D "shields". Initial work has already been conducted on current transformers and a harmonic detector chip has been identified.
A significant aspect of the plan is to connect the wireless system to a cloud computing architecture for storing data in a distributed database. After developing the basic sensing and data transmission systems, we intend to focus on harmonic detection. Higher-frequency harmonic content in voltages and currents, inevitable and unwanted, presents a serious challenge to utility companies trying to minimize waste and control production costs. Furthermore, it interferes with critical circuit protection gear. Gaining an improved understanding of harmonics in power systems will require means to collect large amounts of data. Wireless technology is essential to collecting and manipulating these data.
The first phase of the work, to be performed in the summer of 2012, will be to assemble an ECE227/427 demo board that simulates a simplified power transmission/distribution system. An undergraduate student will work full-time for the summer to assemble the demonstration board and then to link the data collection system to a LabView interface. Once these goals are achieved, work on the wireless components of the system will commence. This project is complex and will certainly extend into the Fall semester. We further expect that it will develop into a senior design project involving several engineering seniors.
Students applying to work on this project must demonstrate a good working knowledge of AC circuits. Project advisors are W. Heinzelman (wendi.heinzelman@rochester.edu) & T. B. Jones (jones@ece.rochester.edu).
Prof. Kevin Parker
Department of Electrical and Computer Engineering
kevin.parker@rochester.edu
Research Project: Medical Ultrasound and Imaging the Elastic Properties of Tissue
Prof. Kevin Parker, Dept. of ECE, BME, and Radiology
This project uses a combination of Ultrasound Doppler Imaging and shear wave analysis to image the hidden elastic properties of tissues. Tissue stiffness and viscosity are two simple and important properties of tissues, yet they cannot be imaged and quantified with normal radiology scans. Instead, special shear wave propagation in tissue, with customized imaging and analysis techniques, have been developed. Applications include the detection of hard cancers in soft tissue (such as the liver, thyroid, and breast), and the staging of liver fibrosis. Undergraduates can assist in the research in a number of ways, as listed below (but no one person is expected to do all of these tasks!)
-data collection, and scanning samples
-data analysis
-preparation of tissue "phantoms" or test objects
-Matlab image processing
-statistical analysis
-Finite Element Analysis of shear wave propagation
-Wave equation solutions
Prof. Robert Buckley
Department of Electrical and Computer Engineering
robert.buckley@rochester.edu
Research Project: Opponent Color Transform of Natural Images
Color image communication systems, whether engineered or the human visual system, typically transform the red, green and blue values produced by image sensors to opponent color signals that are used for transmission and processing. The opponent color signals consist of an achromatic or luminance signal and two color difference or chrominance signals. Since the NTSC standard for color television introduced the YIQ luminance-chrominance signals for color television transmission in the 1950’s, almost all color image systems, including the JPEG, JPEG 2000 and JPEG-XR compression standards, have used some form of opponent color transform. These transforms make various claims to computational efficiency and superior coding gain or visual performance. The goal of this project is to evaluate these transforms with respect to a set of natural images and compare them to the processing in the human visual system and the characteristics of scenes.
