We are reorganizing this page. Some links may not work right now (due to reconfiguration of the web server at College of Engineering). We apologize for the inconvenience.
Continuous Analysis of Many Cameras (i.e., "BigData")
Energy Management of Office Equipment
Mobile + Cloud Computing
Autonomous Information Dissemination in Randomly Deployed Sensor Networks (2007-2012)
A wireless sensor network (WSN) is a wireless network consisting of spatially distributed autonomous devices using sensors to cooperatively monitor physical or environmental conditions, such as temperature, sound, vibration, pressure, motion or pollutants, at different locations. When sensors are far from the receiver, individual sensor or hop-by-hop transmission doesn't have enough energy. This project is to create a new sensor network architecture, AIDA, that uses antenna arrays of the distributed sensor nodes to create directional radio waves which can give stronger signals than individual antenna of the sensor nodes. Because of wave reinforcement and cancellation, the signals need to be highly directional and in phase at the receiver point.
Energy Conservation for Multimedia Systems on Mobile Devices (2007-2011)
Multimedia content such as image and video usually require memory- and computation-intensive processing and thus cost a large amount of energy. In this project, we devise energy-efficient implementations of several multimedia applications on mobile device. We also explore the server-client communication and computation cost to make offloading decisions. The ultimate goal is to develop a general and efficient framework for multimedia applications on mobile devices.
Video: Energy Conservation by Adaptive Feature Loading for Mobile Content-Based Image Retrieval
Resource Driven Transmission and Display of Multimedia in Mobile Devices (2005-2010)
There has been a dramatic growth in the usage of images and videos in mobile devices for fields like personal communication, surveillance and tracking, and health-care. Since mobile devices are battery-operated, energy conservation is very important. We develop techniques for transmission and display of multimedia for devices with different amounts of resources. We remove fine details from images and videos, while preserving significant structural information. Removal of fine details reduces the sizes of images and videos, thereby reducing their transmission energy. We also develop layout generation techniques for adapting multimedia presentations for devices with different display resolutions.
Rendering Source Code:
Privacy-Perserving Computation Offloading (2010-2011)
Offloading allows mobile systems to send heavy computation to high-performance servers and save mobile systems' energy. However, private information becomes available at the servers. This project develops technologies that can protect privacy in offloading.
Mobile Applications on the Cloud (2010-2011)
Combine the advantages of mobility and the resources of cloud computing.
Task Partition for Elastic Computing (2010-2011)
Develop partition algorithms that can simultaneously determine how many cores to use and which tasks should run on each core.
Homogeneous Architecture for Power Policy Integration (2003-2008)
The Homogeneous Architecture for Power Policy Integration (HAPPI) is a software framework that allows the operating system to choose the best power management policy from a library of available policies at run-time. This technique, called automatic policy selection, is important because different policies are necessary to achieve maximum energy savings for different workloads and devices. HAPPI is implemented in the Linux 2.6.17 kernel. The source code is available under the GNU Public License version 2.0.
Video: A Homogeneous Architecture for Power Policy Integration.
Cross-Layer Energy Management by Operating Systems and Applications (2004-2008)
Energy conservation has become a major challenge in system design. Most existing techniques focus on individual layers: hardware, operating system (OS), and applications. This project focuses on OS and applications and we propose a cross-layer approach to help OS and applications collaborate for better power management. We instrument OS to measurement power and follow a set of accounting rules to determine the amount of energy consumed by each process. These processes can adjust their behavior to use energy efficiently. For portable systems powered by batteries, some processes may off-load computation to a remote server that is grid-powered. A programming environment has been developed to assist programmers in improving energy efficiency. This project has impacts on improving the energy efficiency of electronic systems to provide services for longer periods of time for mobile users.
Video: A Programming Environment with Runtime Energy Characterization for Energy-Aware Applications.
Source code: Energy-efficient Programming ENvironment (EPEN)
Energy Efficient Mobile Robots (2003-2007)
This project studies energy conservation through the coordination of multiple distributed and autonomous mobile robots. A set of algorithms have been developed for motion planning, deployment strategy, and replacement of sensor nodes.
Joint Power Management in Storage (2002-2006)
This project considers power management in the storage hierarchy of computers, especially the interactions between memory and disks. Power management in memory affects the access patterns and idleness of disks. Hence, memory and disks should be managed together to save energy.
Sensor Network: Detect, Optimize, and Recover (SeNDOR) (2004-2007)
The SeNDOR project focuses on designing low-overhead, high-performance distributed monitoring techniques in order to debug and analyze performance during development and deployment of large scale wireless sensor networks. This work focuses on creating tools to insert monitoring code within applications and extend this monitoring framework to areas such as security.