By looking at the latest electronic communication devices that have emerged over the past few years, it's clear that the trend of smaller, portable devices is strong and expected to continue. Yet while all these notebooks, netbooks, and tablet PCs are becoming more and more popular, their explosive growth also poses a problem: these wireless devices are hogging the already congested lower microwave frequency region of the wireless spectrum.
This congestion problem was not unanticipated by electrical engineers, who, for the past two decades, have been developing new wireless technologies that use different parts of the electromagnetic spectrum. Specifically, these wireless technologies are exploiting the large, unused bandwidths of extremely high frequency (EHF) microwaves in the millimeter-wave (mm-wave) frequency region. One particular area of interest is the unlicensed 60 GHz frequency band, which has 5-mm wavelengths. (In contrast, the heavily burdened lower microwave regions have frequencies of 2-4 GHz, corresponding to wavelengths of 7.5-15 cm.)

However, the 60 GHz frequency band is not without challenges, either. Since wireless signals at 60 GHz frequencies have inherently high propagation losses, they are targeted toward short-range, in-building, high-speed applications. To maintain strong incoming wireless signals for buildings, many antenna base stations must be built near customers. These base stations, in turn, would receive broadband signals from a smaller number of distant central offices. The signals between central offices and base stations would be transmitted through long-range optical fibers. Since such a system uses both optical fibers and mm-wave wireless transmission, the technology is called “fiber-wireless” (Fi-Wi).

The advantage of bimodal Fi-Wi systems is that they can enjoy the strengths of both optical and wireless technologies - specifically, the inherently large bandwidth of optical fiber and the large, unused bandwidth in the mm-wave wireless spectrum. For this reason, a hybrid system has the potential to provide very high data transmission rates with minimal time delay.

Recently, a team of electrical engineers working on fiber-wireless technologies has analyzed the progress made in this field over the past two decades. In a paper published in the Journal of Lightwave Technology, Christina Lim, from the University of Melbourne, and her coauthors have presented an overview of the many different techniques proposed to optically transport mm-wave wireless signals and overcome some of the challenges involved.

 

RSA authentication is a popular encryption method used in media players, laptop computers, smartphones, servers and other devices. Retailers and banks also depend on it to ensure the safety of their customers' information online.

The scientists found they could foil the security system by varying the voltage supply to the holder of the "private key," which would be the consumer's device in the case of copy protection and the retailer or bank in the case of Internet communication. It is highly unlikely that a hacker could use this approach on a large institution, the researchers say. These findings would be more likely to concern media companies and mobile device manufacturers, as well as those who use them.

Andrea Pellegrini, a doctoral student in the Department of Electrical Engineering and Computer Science, will present a paper on the research at the upcoming Design, Automation and Test in Europe (DATE) conference in Dresden on March 10.

"The RSA algorithm gives security under the assumption that as long as the private key is private, you can't break in unless you guess it. We've shown that that's not true," said Valeria Bertacco, an associate professor in the Department of Electrical Engineering and Computer Science.

These private keys contain more than 1,000 digits of binary code. To guess a number that large would take longer than the age of the universe, Pellegrini said. Using their voltage tweaking scheme, the U-M researchers were able to extract the private key in approximately 100 hours.

They carefully manipulated the voltage with an inexpensive device built for this purpose. Varying the electric current essentially stresses out the computer and causes it to make small mistakes in its communications with other clients. These faults reveal small pieces of the private key. Once the researchers caused enough faults, they were able to reconstruct the key offline.

This type of attack doesn't damage the device, so no tamper evidence is left.

"RSA authentication is so popular because it was thought to be so secure," said Todd Austin, a professor in the Department of Electrical Engineering and Computer Science. "Our work redefines the level of security it offers. It lowers the safety assurance by a significant amount."

Although this paper only discusses the problem, the professors say they've identified a solution. It's a common cryptographic technique called "salting" that changes the order of the digits in a random way every time the key is requested.

"We've demonstrated that a fault-based attack on the RSA algorithm is possible," Austin said. "Hopefully, this will cause manufacturers to make a few small changes to their implementation of the algorithm. RSA is a good algorithm and I think, ultimately, it will survive this type of attack."

The paper is titled "Fault-based Attack of RSA Authentication." This research is funded by the National Science Foundation and the Gigascale Systems Research Center.