Computer Engineering

Yehia Massoud
Student needs: up to 3
Suggested Elec 599 project:

• Topics in Design Automation
  

Kartik Mohanram
Student needs: 1-2
Suggested Elec 599 project:

• Topics in fault tolerant computing
  

Lin Zhong
Student needs: 3
Suggested Elec 599 project:

• (Joint w/ Frank Tittel)
  Development of Open-Access Networkable Photonic Sensors (PHOTONS) for Security, Industrial, and Environmental Applications
  Ashutosh Sabharwal, Farinaz Koushanfar, Frank Tittel, Gerard Wysocki, and Lin Zhong
  Project Description
  Chemical sensing of trace gases can provide critical information about the world around us. Consider two compelling examples: when measured in parts-per-billion (ppb), detecting trace amounts of specific gases can allow us to pinpoint explosive material in luggage or can diagnose diseases from exhaled breath. Therefore, the development of low-cost, ultra compact sensors capable of measuring trace gases at ppb levels is a timely research opportunity for several novel applications. This project will develop an integrated trace-gas sensing platform to achieve optimum and robust sensor performance, integrated networked platforms and at-scale deployments. The networkable photonic sensors platform (PHOTONS) is built from the ground-up to enable researchers, developers, and commercial organizations to rapidly experiment with never-before-possible trace-gas sensing applications in security, industrial and environmental monitoring in an affordable, portable, and power-efficient manner. The key innovations include advanced sensor technologies, system platform development, and a framework to accelerate development of novel applications. The project will involve the following two related tasks.
  
  
  Sensor Miniaturization: The proposed project will focus on the design novel trace-gas sensors using our recent breakthroughs in quartz-enhanced photoacoustic spectroscopy (QEPAS). The new sensors will possess the required sensitivity, selectivity and time response as well as lower power requirements.
  Networkable System Integration: The project will employ commercial-off-the-shelf discrete hardware components to build necessary networking and energy supply sub-systems to support the miniature sensor development mentioned above. The sensors will then integrate to control sensing accuracy and manage energy consumption at a fine granularity in a coordinated manner. The integration will allow new and useful research for optimizing network lifetime and performance.
  
  
  The project is multidisciplinary and the participating student(s) will interact with a team of ECE faculty members working in laser science, wireless communication, embedded systems, and statistics to ensure its success. Interested students should contact Professor Frank Tittel (fkt@rice.edu) for more information.

• Efficiency-Driven Cross-Layer Design of Mobile Broadband Access
  Performance has been the primary goal in the design of modern high-performance wireless network. As high-performance network technologies, such as 802.11, have achieved far higher data rate (hundreds and thousands of Mbps) than typically required in most everyday mobile applications, their energy efficiency has become a critical design concern, in particular for mobile systems. This project will first examine implementation of modern high-performance wireless interfaces and investigate standard-compliant solutions to dramatically reduce the power consumption of active wireless data transfers and maintains the lowest possible energy consumption per bit. It will further study high-performance wireless network protocol stacks and question how high-performance wireless network can be design differently with efficiency, instead of performance, as the primary goal.

• Infrastructure-Device Cooperation for Scalable, Reliable, and Energy-Efficient Mobile Multimedia
  Mobile multimedia applications, including mobile TV, are increasingly popular in mobile phone users. However, they impose considerable challenges in both the content delivery infrastructure (ISP and ICP) and the end devices. For the infrastructure, scalability and reliability are critical for delivering services to a large number of subscribers with guaranteed quality of services, particularly under the situation of massive interest in certain media content. For the end devices, the battery lifetime has already become the bottleneck for mobile multimedia consumption. The state of the art mobile handsets can only afford a couple of hours of mobile TV even without any other usage. Our user studies showed that mobile users are reluctant primarily because of the battery lifetime concern. While there is a clear boundary between the infrastructure and end devices in existing mobile TV solutions, this project will address these challenges with a cooperative approach. The objective is to devise infrastructure mechanisms for end devices to make tradeoffs between the quality of services and battery lifetime as well as explore end device mechanisms to relieve the scalability and reliability burdens on the infrastructure.

Multimedia Communications

Joseph Cavallaro
Student needs: up to 2 students
Suggested Elec 599 projects:

• Algorithms and Architectures for Wireless Receivers:
  Wireless communication systems are now using multiple antennas and orthogonal frequency division multiplexing to achieve higher data rates. Advances in VLSI signal processing architectures are needed to reduce the power consumption and improve the bit error rate performance of these algorithms. New detection and decoding algorithms are being proposed, including so-called sphere detection with LDPC decoding. In this project, we will study the performance tradeoffs and the importance of channel estimation approaches for iterative receivers and the resulting architecture complexities. This project will utilize the testbed resources in the Rice CMC Lab.

• Architecture Analysis of Processors for Wireless Systems:
  Rapid prototyping of architectures for wireless communication systems enable the design of configurable datapath accelerators for signal processing algorithms. In the Rice CMC Lab, System Generator is used to target configurable FPGA devices. These configurable datapaths are typically interconnected with a programmable core processor in a system on chip. A number of recent programmable multi-core architectures have emerged from TI, ARM, and Freescale that may be applicable to next generation wireless systems. In this project, the goal will be to benchmark wireless receiver algorithms to analyze time, area, and power requirements, and to compare the efficiencies of programmable systems with accelerator based systems. Intermediate versions will be used with the FPGA testbed in the CMC lab. The compiler and CAD design tools to be studied will interface with TI, ARM, and Synopsys/Cadence/Mentor design flows.

Networks

Dave Johnson
Student needs: 1 (Keyvan Amiri)
Suggested Elec 599 project:

• COMPASS sensor networking project

Edward Knightly
Student needs: up to 2
Suggested Elec 599 project:

• Topics in analysis and experimental research in multi-hop wireless networks
  

Signal and Image Processing

Richard Baraniuk
Student needs: 2-3
Suggested Elec 599 project:

• Theory and applications fo Compressive Sensing with applictaions to analog-to-information conversion, sensor networks, and non-local means.

Erzsébet Merényi
Student needs: 1-2 (Brian Bue)
Suggested Elec 599 project:

• Topics in neural information porcessing: learning high-dimensional manifolds with self-organized learning and other neural approaches; clustering, classification, prediction, modeling in high-dimensional data spaces; applications in hyperspectral imaging (Earth and space science, planetary surfaces, tissue imaging), and bioinformatics.

599 Home

Comments and questions to: www-ece@ece.rice.edu

Last modified: December 3, 2007