The Community for Technology Leaders


Pages: pp. 20-23

UN Internet Summit Ends in Controversy, Disagreement

A recent United Nations meeting on Internet governance ended in heated arguments, a lack of consensus, and some countries' representatives walking out in protest. The main issue of dispute—largely between democratic and nondemocratic countries—was whether governments should have more control over the Internet.

The World Conference on International Telecommunications (WCIT) in Dubai, United Arab Emirates, was held to discuss and potentially make decisions on the future of telecommunications. However, 12 days of largely closed-door negotiations involving about 1,500 diplomats from 151 countries yielded no general agreement.

Initially, conference sponsors claimed WCIT wouldn't really be about the Internet. However, various proposals brought up during the conference sought to nationalize some governance-related activities, such as domain naming and IP address numbering, and would have obliged countries to take more steps regarding online security.

A key disagreement thus became whether the International Telecommunications Regulations treaty, adopted in 1988, would be extended to address the Internet. This could have let the Telecommunication Standardization Sector of the UN's International Telecommunication Union (ITU) direct many aspects of Internet governance, which various international organizations—such as the Internet Engineering Task Force—currently handle.

Opponents of extending the treaty expressed concern that this would let authoritarian governments exert increasing control over the Internet. They also said such control would be unacceptable because it would change the effective model by which policy has been made not by governments but by groups of Internet users and experts.

In another point of dispute, Algeria, China, and Iran objected to a US-backed proposal to include human rights obligations in a new telecommunications treaty, saying they didn't belong in such an agreement.

At one point during WCIT discussions, representatives from Canada, the UK, the US, and various other nations walked out to protest attempts by some countries to expand governmental Internet control.

Ultimately, the 89 countries favoring such increased control—including Iran, Russia, and some other nations that already practice Internet censorship—adopted a resolution to that effect. However, it is not binding on the 55 opposing nations, which will continue to follow the 1988 treaty. Thus, there will now be two sets of countries following different Internet-governance rules.

Additional international Internet-governance events are scheduled for this year, such as the World Telecommunication Policy Forum in May.

Carnegie Mellon University Wins $1.17 Billion Patent Lawsuit against Chipmaker

A US federal jury has ruled that Marvell Semiconductor must pay $1.17 billion in damages for infringing two patents related to hard-drive-control chips that Carnegie Mellon University owns.

Marvell could face triple damages because the jury—following a four-week federal court trial—said the company knew it was using the CMU patents without a license.

If upheld, the $1,169,140,271 in damages—the full amount CMU requested—would be the biggest patent verdict in history. Earlier this year, a jury ordered Samsung to pay $1.05 billion for infringing on Apple's intellectual property. There have been larger patent verdicts, but none held up after appeal or judicial review.

Marvell has said it will appeal the recent ruling and is also waiting for a decision on its motion for a mistrial, made before the case went to the jury.

If the verdict holds up, it would be more than the $900 million in profits Marvell earned in all of 2011. Immediately following the jury ruling, the value of the company's stock dropped about 10 percent.

The CMU patents in question involve technology that increases the speed and precision by which circuits in hard drives read data from magnetic disks. The school never started a firm or awarded licenses to a company to manufacture products based on the patents.

During the trial, Marvell contended that CMU didn't invent anything new. The company said Seagate Technology filed a patent that addressed everything in CMU's proposal 14 months before the university submitted its proposal. CMU denied this and said Marvell clearly used its protected technology without a license.

Marvell reportedly sold 2.34 billion hard drives that infringed CMU's patents between 2003 and 2012.

This case is the latest example of how universities are taking to court companies that allegedly use their patents without permission. Recently, Cornell University sued Hewlett-Packard over a processor patent, and Northeastern University filed a suit against Google over a Web patent.

Researchers Develop First Flexible, Transparent Image Sensor

Two Austrian researchers have developed the world's first flexible, transparent image sensor, which could be used in many places for which traditional image sensors are unsuitable. This would enable the creation of imaging-producing or sensing applications that could be embedded in everyday objects such as car windshields.

The technology could also be used in user interfaces that replace touchscreens. The approach's ability to sense a person's movements and generate signals would eliminate the need for users to actually touch the screen.

Professor Oliver Bimber and PhD student Alexander Koppelhuber of Johannes Kepler University of Linz's Institute of Computer Graphics, working with Microsoft Research in the UK, developed the new sensor and have already built the first prototypes.

The biggest challenge in developing flexible, transparent image sensors has been the accurate reconstruction of images on the sensor's surface. Bimber's team addressed this issue by doping a luminance-concentrator thin film with fluorescent dye particles. When an image focuses on the film, the substance absorbs light wavelengths, re-emits them at a lower frequency, and transports them to the film's edges.


Figure    Two scientists have developed a flexible, transparent image sensor that could be used on car windshields and other places where such sensors currently can't be deployed.

Using photo sensors at the edges and an optical technique the researchers aren't yet discussing in detail, the sensor can measure the distribution of the 2D light field that travels inside the film. The system uses this information, as well as mathematical techniques the Kepler researchers developed, to quickly perform in parallel the complex calculations necessary to reconstruct the image on the film's surface.

The new sensor's lightweight, flexible, clear film could scale as desired and would be inexpensive to manufacture.

The sensors could also be layered. "This allows colors to be imaged—in underlying pixels—one color per layer," Bimber explained. The depth of the pixel presentation would improve the resolution. "Conventional image sensors capture colors in pixels alongside one other, reducing effective resolution," Bimber said.

The layering would also enable presentations with a high-dynamic range between the lightest and darkest parts of an image.

"Currently, we're working on increasing the image quality and resolution," Bimber noted.

Vendor Works on Secure OS for Industrial Control Systems

Security vendor Kaspersky Lab is designing a secure operating system for the industrial control systems (ICSs) that help control nuclear-power plants, utility infrastructure operations, factories, and other important facilities.

In recent years, ICSs have proven to be vulnerable to cyberattacks, which could have serious and deadly consequences such as equipment malfunctions and the interruption of public services. Hackers have already targeted the systems with high-profile malware packages including Stuxnet, Duqu, Flame, and Gauss.

Experts say most ICSs were built many years ago to maintain operations at all times, not with security in mind. At the time, network-based attacks weren't common enough to worry about. Most data protocols in ICSs don't even require user identification or authorization, according to Kaspersky.

The systems are thus highly vulnerable to today's sophisticated hackers, including those employed by national governments, which have considerable resources to spend on developing attacks.

Rewriting all ICS software would be highly complicated, expensive, and time-consuming. Instead, Kaspersky has proposed building a secure OS on which an organization could run existing ICSs.

Kaspersky says it plans to build its secure operating system from scratch, with assistance from ICS vendors and users. A key, the company states, is verifying that the OS won't permit either vulnerabilities or code for legitimate purposes that hackers could exploit.

To enhance safety, the OS kernel will have to be very efficient and contain as little code as possible. The firm declined to provide more details, citing the importance of confidentiality to the project.

New Approach Would Modernize Today's Internet

The Internet is increasingly used for datacenter operations, as well as for mobile computing and communications. However, it wasn't designed for such purposes.

Now, Princeton University researchers—led by assistant professor Michael Freedman—have developed a technology designed to make the Internet better for such applications. Their open source Serval system is an adapted Internet architecture designed to support today's dynamic network environments in which the ability to access services is critical.

The Internet was designed when users connected to specific computers in specific locations. Now, though, users connect more often to services such as Google. This can be difficult because IP addresses are still linked to specific machines.

Having users move between wireless networks or having companies move traffic among servers requires increasingly complex and expensive workarounds. To avoid this, Serval adds a new layer of information called the service access layer to some packets in a transmission. This layer identifies a service so that, if an IP address changes, an application still gets the information it needs to keep working with the service.

Serval's approach is similar to that of software-defined networking (SDN), which moves control operations from hardware to software. However, the Princeton researchers say, Serval is focused on the relationship of high-level services—such as user mobility and virtual machine migration—to the network.

SDN focuses on operations within the network. Thus, a mobile phone user could access and stay connected to a Web service while walking or driving around. Or users could watch videos while their laptop switches between Wi-Fi and cellular services.

The Princeton scientists have started a small Serval demonstration network and are beginning to test their technology in larger and more complex systems. They still must determine how Serval will function in large deployments and over the entire Internet.

The US National Science Foundation, DARPA, the US Office of Naval Research, and networking equipment vendor Cisco Systems are funding the Princeton project.

Tiny Wireless Computer Tracks Patients' Health

An academic researcher has developed a tiny wireless computer that patients could wear to monitor their health, determine whether they're taking their medication as required, or identify whether they're in danger of falling. This technology reflects a growing trend in which miniature devices are used in medical settings.

University of Texas at Dallas assistant professor Roozbeh Jafari's new system is about the size of a clothing button, making it easily wearable. A key to making the system so small was eliminating the need for bulky batteries and designing it to use power efficiently. For example, Jafari's device runs only the most important data needed for its designated task, thereby avoiding unnecessary energy-consuming activities. In addition, the sensors could be optimized to use different amounts of energy for different health-monitoring purposes.

The machine doesn't need to run a lot of information all the time, Jafari said, because "signals and events observed from the human body tend to change slowly. The human body's physics and kinematics reduce the likelihood of random body signals and movements."

The system includes biosensors that collect health-related data from a wearer, as well as microcontrollers that process and communicate the information.

Small health monitors will be increasingly important in the future, said Jafari, director of UT Dallas' Embedded Systems and Signal Processing Lab. "Growing demand for healthcare-monitoring applications will require students, engineers, and healthcare professionals to design, develop, deploy and operate [such] wearable systems," he explained.

For work on his technology, Jafari received a 2012 US National Science Foundation Faculty Early Career Development award given to junior faculty members who, according to the NSF, "exemplify the role of teacher-scholars through outstanding research, excellent education and the integration of education and research within the context of the mission of their organizations."


Figure    People could wear a button-sized wireless computer to monitor their health, including their vital signs and whether they are taking medication as directed.

64 ms
(Ver 3.x)