Award Abstract #0521227
MRI:Development of Energy-Efficient Embedded Systems for Wireless Sensor Networks

NSF Org:	CNS Division of Computer and Network Systems
Initial Amendment Date:	August 4, 2005
Latest Amendment Date:	August 4, 2005
Award Number:	0521227
Award Instrument:	Standard Grant
Program Manager:	Rita V. Rodriguez CNS Division of Computer and Network Systems CSE Directorate for Computer & Information Science & Engineering
Start Date:	August 1, 2005
Expires:	July 31, 2009 (Estimated)
Awarded Amount to Date:	$400000
Investigator(s):	Anthony Ephremides tony@eng.umd.edu (Principal Investigator) Bruce Jacob (Co-Principal Investigator) Rajeev Barua (Co-Principal Investigator) Gang Qu (Co-Principal Investigator) Peter Petrov (Co-Principal Investigator)
Sponsor:	University of Maryland College Park 3112 LEE BLDG COLLEGE PARK, MD 20742 301/405-6269
NSF Program(s):	UNDISTRIBUTED PANEL/IPA FUNDS
Field Application(s):	0000912 Computer Science
Program Reference Code(s):	HPCC, 9218, 9135
Program Element Code(s):	9199

ABSTRACT

This project, developing experimental prototypes of intelligent energy-efficient embedded systems for use in nodes of sensor networks, plans to acquire the hardware equipment that

-Builds energy efficient embedded sensor systems;

-Characterizes their energy consumption parameters; and

-Verifies and evaluates new designs.

The goal is to translate ongoing research on energy efficiency to development of advanced sensors that incorporates as many energy-saving features as possible while at the same time remain light, small, inexpensive, and relatively flexible. The investigation covers the spectrum: energy-efficient circuitry, energy-aware architectures, energy-aware embedded systems compilers, energy-efficient system integration, and energy-efficient communication and networking. The research needs to

-Compare analog and digital circuit implementations, Modify standard processor architectures,

-Develop dynamic memory management schemes, Integrate the components, and

-Tailor those (chosen for overall efficiency) to carefully chosen algorithms for processing and transmission

Virtually all wireless networks are crucially dependent on portable finite energy sources. In particular, sensor networks for a large number of applications (such as event detection, monitoring, and data collection) depend on non-renewable energy sources. Both dominant forms of energy used in sensor networks, RF Transmission and Processing, exhibit expenditure modes that depend on a large number of factors, from the bottom layer up to the network protocol stack. Clearly, there is a fundamental tradeoff between energy spent at transmission and energy spent for processing. For example, data collected by sensors can undergo complex compression that reduces the number of transmitted bits and hence saves transmission energy at the expense of increases processing energy; thus the component of the sensor that includes the communication system must also be optimized with respect to energy efficiency. Hence, this work experiments with different scenarios of network communication. Through both simulation and prototype implementation, two particular issues related to communication and networking are studied: To

-Integrate communication functions such as modulation and coding with hardware design (e.g., power amplifier)

-Evaluate power consumption of transmission vs. processing and longer vs. shorter hops with actual physical tests

-To calculate the energy consumption tradeoffs, measurements and display equipment are necessary. Specific research projects include

-Energy-efficient circuit design; Micro-controller architectures; Memory hierarchies;

-Component integration, and Communication protocols.

Broader Impact: The project should have a huge educational impact. While training the next generation of researchers, this work contributes a freely available tool for cycle-level simulation to accurately estimate energy consumption of the design. The need for deployment of long-lived sensor networks is well documented and affects directly the commercial and military sectors. An opportunity exists for significant reductions in energy consumption.

PUBLICATIONS PRODUCED AS A RESULT OF THIS RESEARCH

(Showing: 1 - 3 of 3).

Leeger Yu, Anthony Ephremides.  "Enhancing Energy Efficiency for Data Fusion in Sensor Networks with a Detection Mission,"  Proceedings of the IEEE WPMC,  2008, 

Nghi Nguyen, Angel Dominguez, Rajeev Barua.  "Mempry Allocation for Embedded Systems with a Compile-Time Unknown Scratch-Pad Size,"  ACM International Conference on CASES,  2005, 

Nghi Nguyen, Angel Dominguez, Rajeev Barua.  "Scratch-Pad Memory Allocation without Compiler Supports for java Applications,"  Proc. of the ACM International Conference on Computing Systems,  2007,  p. 1.


(Showing: 1 - 3 of 3).

Please report errors in award information by writing to: awardsearch@nsf.gov.