March
2005
Professor Ali Niknejad's other research focuses on the development of inexpensive radios for indoor wireless networks with ten times the bandwidth of conventional WiFi technology. |
In UC Berkeley electrical engineer Ali Niknejad's imagination, tomorrow's personal computers will look more like iPods than laptop PCs. You'd carry all of your data in your pocket. Need a monitor? Walk up to a wall screen and forge an instant wireless link. Inside your office, the device might connect to the local WiFi network for online access. Step outside and the Internet connection seamlessly switches to cellular. Such a device would require either several radios capable of operating at different frequencies, or a single radio with a bit of brains and a lot of flexibility. Niknejad is working on the latter.
"We're trying to build a cognitive radio that would be aware of its environment," says Niknejad, a member of the Berkeley Wireless Research Center. "It would decide what specifications like frequency and power consumption are appropriate and adjust itself accordingly."
This integrated circuit, a highly-tunable wideband voltage-controlled oscillator, was designed and tested by graduate student Axel Berny, It will form the core of a universal frequency synthesizer chip for the cognizant radio. [View larger image] |
The wireless world -- from Bluetooth mobile phone headsets to 3G networks -- is a mess of myriad technical specifications and standards. For example, most cellular phones in the United States operate in the 1-2 GHz frequency range. To complicate matters, a single cellular standard may be implemented in different bands depending on the country. As anyone who has traveled from the US to Europe knows, this makes global connectivity a big challenge unless you have a high-end phone loaded with components that can change from band to band. On the shorter range, wireless local area networks and Bluetooth devices occupy the 2-3 GHz bands while long-awaited ultra-wideband (UWB) promises high-speed delivery of multimedia content at still another frequency.
"You really need to have one device that communicates with a plethora of standards," Niknejad.
Not only would this enable single devices to access many wireless networks, the flexibility could keep new products from becoming outdated even before they hit the marketplace.
"By the time committees come up with a standard, and industry designs radios that meet the specifications, the technology is often obsolete," Niknejad says. "For example, we might use one modulation scheme today but in a year we'll probably find a better one."
|
|
The Cognizant Universal Radio (COGUR) that Niknejad is developing with graduate student Axel Berny and others was built from the bottom up to be flexible. The COGUR design combines analog and digital components into a single "programmable" system, Niknejad explains.
"We're figuring out how to put knobs in the building blocks of radio circuits," he says.
Already, the researchers have fabricated a key component for COGUR. A voltage-controlled oscillator (VCO) regulates the radio waves that carry a signal. The frequency of a traditional VCO can be tuned by about 10 to 20 percent up or down, but Niknejad and Berny's device is capable of 100 percent variation. For example, it can shift frequency from one gigahertz to two, or five gigahertz to ten. By adding a frequency synthesizer that Berny is now testing, "it could hit any frequency you want," Niknejad says.
Not only is the frequency programmable, but other variables are also tunable on the fly. Take power, for instance. A Bluetooth signal need only travel a few feet, requiring far less juice than, say, a cell phone conversation.
"A cognitive radio will find the optimal current level to get the best performance without wasting power," Niknejad says.
Right now, the researchers are designing the various building blocks that they someday hope to combine into a fully functional cognizant radio. In the next few years, the goal is to integrate all of the components on a single piece of silicon that would not only make our wireless devices smarter and more powerful, but also smaller and cheaper.
"We're going for the holy grail here, putting an entire cognizant radio onto a single chip," Niknejad says. "Today's technology is almost there."
Ethanol Stirs Eco-Debate
by David Pescovitz
Tad Patzek is professor in the Department of Civil and Environmental Engineering. |
In 2004, approximately 3.57 billion gallons of ethanol were used as a gas additive in the United States, according to the Renewable Fuels Association (RFA). During the February State of the Union address, President George Bush urged Congress to pass an energy bill that would pump up the amount to 5 billion gallons by 2012. UC Berkeley geoengineering professor Tad W. Patzek thinks that's a very bad idea.
For two years, Patzek has analyzed the environmental ramifications of ethanol, a renewable fuel that many believe could significantly reduce our dependence on petroleum-based fossil fuels. According to Patzek though, ethanol may do more harm than good.
"In terms of renewable fuels, ethanol is the worst solution," Patzek says. "It has the highest energy cost with the least benefit."
Ethanol is produced by fermenting renewable crops like corn or sugarcane. It may sound green, Patzek says, but that's because many scientists are not looking at the whole picture. According to his research, more fossil energy is used to produce ethanol than the energy contained within it.
Patzek's ethanol critique began during a freshman seminar he taught in which he and his students calculated the energy balance of the biofuel. Taking into account the energy required to grow the corn and convert it into ethanol, they determined that burning the biofuel as a gasoline additive actually results in a net energy loss of 65 percent. Later, Patzek says he realized the loss is much more than that even.
"Limiting yourself to the energy balance, and within that balance, just the fossil fuel used, is just scraping the surface of the problem," he says. "Corn is not 'free energy.'"
Recently, Patzek published a fifty-page study on the subject in the journal Critical Reviews in Plant Science. This time, he factored in the myriad energy inputs required by industrial agriculture, from the amount of fuel used to produce fertilizers and corn seeds to the transportation and wastewater disposal costs. All told, he believes that the cumulative energy consumed in corn farming and ethanol production is six times greater than what the end product provides your car engine in terms of power.
Patzek is also concerned about the sustainability of industrial farming in developing nations where surgarcane and trees are grown as feedstock for ethanol and other biofuels. Using United Nations data, he examined the production cycles of plantations hundreds of billions of tons of raw material.
"One farm for the local village probably makes sense," he says. "But if you have a 100,000 acre plantation exporting biomass on contract to Europe , that's a completely different story. From one square meter of land, you can get roughly one watt of energy. The price you pay is that in Brazil alone you annually damage a jungle the size of Greece ."
If ethanol is as much of an environmental Trojan horse as Patzek's data suggests, what is the solution? The researcher sees several possibilities, all of which can be explored in tandem. First, he says, is to divert funds earmarked for ethanol to improve the efficiency of fuel cells and hybrid electric cars.
"Can engineers double the mileage of these cars?" he asks. "If so, we can cut down the petroleum consumption in the US by one-third."
For generating electricity on the grid, Patzek's "favorite renewable energy" to replace coal is solar. Unfortunately, he says that solar cell technology is still too immature for use in large power stations. Until it's ready for prime time, he has a suggestion that could raise even more controversy than his criticisms of ethanol additives.
"I've come to the conclusion that if we're smart about it, nuclear power plants may be the lesser of the evils when we compare them with coal-fired plants and their impact on global warming," he says. "We're going to pay now or later. The question is what's the smallest price we'll have to pay?"
Professor Luke Lee is also director of Biomolecular Nanotechnology Center and co-director of the Berkeley Sensor & Actuator Center. (Peg Skorpinski photo) |
During the 1970s, UC Berkeley pioneered the tools and techniques that enabled hundreds of thousands of components to be packed onto a tiny computer chip. Thirty-years later, a similar revolution is taking place on campus. Bioengineering professor Luke Lee and graduate students Paul Hung and Philip Lee are developing integrated circuits for biology rather than bits. The new technology could lead to automated "laboratories" the size of fingernails that accelerate drug discovery, synthetic biology, stem cell research, and the development of new biomaterials for implants.
Rather than use transistors to switch voltages racing through wires, these laboratories-on-a-chip employ microscopic valves, pumps, cellular manipulators, and other tiny components to test hundreds of biological samples in parallel. Lee's latest creation shrinks a workhorse of cell biology, the standard petri dish, so that 100 of them fit into a two-millimeter square device. It's a tour-de-force in microfludics, hair-thin plumbing systems capable of transporting nanoliter volumes. (A nanoliter is one-billionth of a liter.)
"The array could enable 100 different cell-based experiments to run in parallel," he says. "For example, a biopharmaceutical company might watch how tumor cells respond to various concentrations of a drug to help determine the correct dosage for the desired response. It's true quantitative biology."
Photograph of the microfluidic cell culture array. A 10 x 10 array of microchambers was fabricated on a 2 x 2 cm device. The port at the left provided continuous perfusion of medium uniformly across the array. The port at the right was the outlet for the medium. Reagents and cells were loaded from the top and flow out through the bottom port. (courtesy the researchers) |
To use Lee's device, several cells are loaded into each of the millimeter-sized chambers riddling the plastic chip. Nutrients are continuously pumped through the array while waste flows out. Various kinds of chemicals, a new drug for example, can then be injected in highly specific amounts into each chamber. Eventually, the chip might be outfitted with Lee's nanoscope that would automatically analyze the biochemical reactions. Right now though, he says, desktop microscopes are good enough for scientists to observe how the cells differentiate or react to the infusions.
"Our initial aim is just to speed up currently-available bioassays," he says.
A scanning electron micrograph photo of a single microfluidic culture unit before bonding to a coverglass. Multiple perfusion channels surround the main culture chamber. The microchamber has just 1 mm in diameter. (courtesy the researchers) |
Today, these kinds of experiments involve growing cells in standard Petri dishes and then adding chemicals to the mix using pipettes. Not only will the microfluidic approach be much faster and precise, Lee explains, but "the continuous flow of nutrients through the channel makes feeding more uniform, similar to our own physiological conditions."
Along with dramatically improving drug development and screening, the high-throughput cell culture array promises to be a godsend for bioengineers developing new biologically-compatible materials for medical implants and other in vivo applications. For instance, biomaterials that are candidates for the construction of artificial bones or heart valves could be introduced into Lee's chip by the dozens. Studying their impact on the cells could then aid in the identification of the most biologically compatible material.
Currently, Lee is designing his next-generation large scale microfluidic cell culture array. The new model is based on a 16 x 24 array of chambers, enabling nearly 400 separate assays to be conducted in one shot. Meanwhile, Lee's developing other devices in his arsenal of Biologic Application Specific Integrated Circuits (BASICs), named for ASICs, computer circuits design, customized, or programmed for a specific application. The goal of BASICs, Lee says, is to create a standardized library of tools for quantitative systems biology.
For example, one device that measures how cells communicate might provide insight into Alzheimer's disease or someday inform the engineering of a bioartificial retina. Already, Lee is collaborating on such efforts with synthetic biologists in Berkeley 's Department of Molecular and Cell Biology and College of Chemistry and the University of California , San Francisco . In the future, Lee says, multiple BASICs might be microfluidically linked to manipulate and prepare biological samples for a variety of "on-chip" assays.
"Berkeley electrical engineers led the way in computer chips," Lee says. "Now we have the opportunity to distinguish Berkeley as the leader in new biological labs-on-chips too."
ENGINEERING HUMOR : NE alum Darren Bleuel spends nine to 10 hours a week writing and drawing the Nukees comic strip. |
One day in 1996, then NE Ph.D. student Darren Bleuel (B.S. '93 Ph.D. '03 NE) was hanging out with his NE friends, reading the Daily Cal funnies. "These comics suck," one friend declared. Then he turned to Bleuel and said, "You should do a strip."
"About what?" asked Bleuel.
"About us," came the reply. "About us Nukees."
Ah, Bleuel knew, the material was just waiting to be mined. Comedy born from engineering students slaving away in the industrial environs of Etcheverry Hall, bent double under the demands of professors, pondering the nature of science over beers, navigating a minefield of personal relationships, and oh the characters, the gloriously weird characters! Bleuel didn't hesitate. A comic strip named "Nukees" -- now that would be funny. Only one problem: He couldn't draw.
Any engineer knows a problem only exists to be worked down to its solution. Bleuel didn't know how to draw, but he could copy with the best of them, so copy he did: Peanuts, Calvin and Hobbes, anything he could get his hands on.
"I got good enough so I could eventually teach myself to draw," he says. In January 1997, Bleuel published his first Nukees strip in the Daily Cal, and the "atomic comic" was born.
"On the outside, it's about engineers and engineering," explains Bleuel. "But it's really about people's feelings and whatever's in my head."
Nukees often takes place in Etcheverry Hall or the southside bar Blakes (Flakes in the strip), where Bleuel sometimes creates the cartoon. It features a whole cast of NE student characters, vaguely derived from Bleuel's NE friends. The characters include, says Bleuel, the main character, Gav, who is "an exasperated cynic-turned-mad-scientist" (and loosely based on its creator), and a "giant, nuclear-powered, robot ant."
Some of the comic situations are culled from Bleuel's own experiences, such as living in Etcheverry for a month, harboring crushes on bartenders, and showing up at the wrong time for a midterm.
After he graduated, Bleuel got a job at Lawrence Berkeley National Lab and is now a health physicist there. He's also the co-owner of Keenspot, a Web site devoted to publishing web comics, including Nukees.
With an eight-year run, the comic strip has hit its stride, and Bleuel says he's happy to keep creating it. "It's therapeutic," he says. "It's nice to get all your feelings down."
Read the strip and learn more at http://www.nukees.com/ .
