|
|
Research
Summaries
New Wireless Communications Technology Advances Multiple
Antenna Systems
Researchers at the UCLA Henry Samueli School of Engineering and
Applied Science have developed a revolutionary integrated circuits
chip for wireless communications that could lead to more reliable
broadband Internet connections and crisper cellular phone calls.
Electrical engineering professor Babak
Daneshrad and Jingming Wang (PhD ’05) have designed a very large
scale integration (VLSI) chip capable of meeting the tremendous
processing power demands required for successful MiMo communications.
Multiple input, multiple output (or MiMo) technology is a communications
technique that uses multiple antennas to send and receive wireless
signals. When received, the combined signals are decoded, allowing
more data to be transmitted without increasing bandwidth requirements.
Wang has designed, developed, and fabricated a VLSI chip capable
of supporting an eight-by-eight MiMo configuration transmitting
a billion bits of information per second, more than 10 times as
much as wireless local area networks (LANS), although it will run
in the same bands.
As it decodes the signals using a matrix inversion operation, the
powerful chip will process 40 to 50 giga operations (gops) per second
for a bandwidth of 10 to 20 megahertz. In comparison, a general
purpose T1 digital signal processor operating 700 megahertz handles
only 1.4 gops per second.
A UCLA testbed will use two of the new chips, each of which will
support the processing computations for 12.5 megahertz of bandwidth
and eight-by-eight antennas.
To read the full story, please click
here.
Revolutionary Software Targets Suspicious Communications
Online
The government’s ability to balance the privacy concerns of lawful
U.S. citizens with effective monitoring of potential terrorists
has proven an increasingly difficult task. But a landmark software
development by researchers at UCLA’s Henry Samueli School of Engineering
and Applied Science may ease some of these privacy concerns by making
the tracking of terrorist communications over the Internet more
efficient, and more targeted, than ever before.
Computer science professor Rafail
Ostrovsky and graduate researcher William Skeith have developed
a new method to mine potential terrorist-related communications
that narrows down the data to only those documents that fit pre-set,
secret criteria chosen by intelligence agencies. The new approach
filters down the information from billions of communications to
just those deemed essential - discarding communications from law-abiding
citizens before it ever reaches the intelligence community.
The truly revolutionary facet of the technology is that it is a
new and powerful example of a piece of code that has been mathematically
proven to be impossible to reverse-engineer. In other words, it
cannot be analyzed to figure out its components, construction, and
inner workings, or reveal what information it’s collecting and what
information it’s discarding. Nor can it be manipulated or turned
against the user.
“Gathering data can be costly and time-consuming for intelligence
agencies. All of the potential data must first be pulled offline
into a trusted and classified environment, and then painstakingly
sifted through,” Ostrovsky said.
“With this new technology, based on highly esoteric mathematics,
the software can be distributed to many machines on the Internet,
not necessarily trusted or highly secure. The software works by
analyzing all of the data and then having the appearance of putting
all the data into a ‘secure box.’ A secret filter inside the box
dismisses some data as useless and collects only relevant data according
to the confidential criteria that can be programmed into the software.
And because it's all done inside encrypted code, it’s not apparent
which, if any, of the data has been selected and kept, except by
the person who has deployed the filter and has the decryption key.”
To read the full story, please click
here.
Harnessing the Power of the Sun for Embedded Sensor Systems
Around the world, solar technologies provide a number of valuable
resources, including light, electricity, and cooling. Now thanks
to breakthrough research at UCLA, solar energy also can power a
class of tiny, environmental sensors.
Scientists studying water resource management in India or tracking
local animal populations are utilizing the solar harvesting system
- Heliomote - developed by researchers in Mani Srivastava’s Networked
and Embedded Systems Laboratory (NESL) that provides unlimited
energy for embedded sensor systems.
Often deployed for long-term studies of specific environmental factors
including light, humidity, and temperature, simple non-mechanical
sensors, or motes, have very low energy requirements and spend most
of their life cycle in “sleep” mode. However, running only intermittently
even the most power-efficient motes will exhaust their batteries
in just over a year.
The harvesting circuit designed by the NESL researchers draws power
from commercial solar panels, manages the use and storage of available
energy, and routes power to the attached sensor.
To enable the system to operate more efficiently, the UCLA team
has developed algorithms that allow a multi-node system to determine
when solar energy is available, and adjust the overall system demands
on any given sensor accordingly.
“Because these are event-driven systems,” said graduate student
Sadaf Zahedi, “we want to maintain a specific energy level in each
of the sensors. So we’ve created a system that can adjust on the
fly and look to the nodes in direct sunlight to run at a higher
duty cycle, limiting the energy demands on nodes at night, or those
located in partial light or shadow.”
Building on their success, the NESL team is exploring other uses
for the solar energy harvesting unit, including medical applications.
To read the full story, please click
here.
Interdisciplinary Team Leads Nanoscale Research in Materials
for Silicon Chips
The integrated circuits in personal computers, cellular phones,
electronic games and other consumer electronics have become significantly
smaller and more densely packed in the last decade - some have nearly
a billion transistors per chip. As the number of transistors on
a single chip increases, the materials used in the transistors also
must be scaled down in size, and researchers are finding that these
elements operate very differently at the nanoscale.
Funded by a $1.3 million grant from the National Science Foundation,
researchers at the UCLA Henry Samueli School of Engineering and
Applied Science are developing methods to strengthen and improve
materials used for interconnect and packaging components for high-tech
chips.
UCLA materials scientist King-Ning
Tu is partnering with mechanical and aerospace engineering professor
Nasr
Ghoniem, who specializes in advanced computer simulations, materials
science and engineering professor Jenn-Ming
Yang, and Nicholas Kioussis, a professor of physics at California
State University, Northridge on three core goals: to strengthen
copper at the nanoscale, improve its reliability, and create a better
insulation material.
Copper, sized at just a fraction of the width of a human hair, is
commonly used for interconnect wires in transistors. When reduced
to only 50 nanometers in width, gravity causes the copper to sag,
creating interference between the wires.
The research team is looking to nano-twinned copper, which is specially
treated to add patterned irregularities, for use at the nanoscale.
The material is 10 times stronger than untreated copper, and loses
none of its electrical conductivity - making it an ideal material
for silicon interconnects.
Advanced computer simulations devised by Ghoniem and Kioussis will
help determine the electromigration tendencies of nano-twinned copper,
helping the researchers to prevent voids or shorts caused by the
movement of atoms.
To read the full story, please click
here. |
|
