Issue No.07 - July (2012 vol.45)
Published by the IEEE Computer Society
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/MC.2012.239
Topics include several cable service providers joining forces to make all of their Wi-Fi hotspots available to customers, researchers tracking taxis via GPS to identify urban transportation problems, a new gesture-control system, hackers injecting advertisements into Wikipedia pages, a faster new Fourier transform version, and an IP-over-xylophone system.
Cable Providers Join Forces on Wi-Fi Hotspots
In an effort to maintain influence in the wireless communications market, five major US cable-service providers have joined forces to let their customers access all of the carriers' Wi-Fi hotspots when traveling outside local networks.
Bright House Networks, Cable-vision's Optimum Online, Comcast's Xfinity, Cox Communications, and Time Warner Cable will make their 50,000 US hotspots available to each paying consumer under the network name CableWiFi.
The participating companies have activated the program in New York City and parts of central Florida, and plan to make it available in Los Angeles, Philadelphia, and parts of New Jersey and Connecticut in the near future. They plan to bring more hotspots online soon, especially in high-traffic places such as malls, parks, beaches, stadiums, and arenas.
Users will have to log in to the hotspots to confirm that they are paying customers. But the companies are developing technology that will store credentials so that users can log in automatically.
The cable providers are responding to stiff competition from high-speed cellular-based data technologies such as Long Term Evolution (LTE), offered by companies like Verizon Wireless.
Cable companies have been trying for years to find ways to attract wireless-data customers. They contend that Wi-Fi is a superior approach for mobile data and will be even more useful when they make additional hotspots available.
CableWiFi expands on existing agreements among Cablevision, Comcast, and Time Warner Cable to share hotspots in the northeast US.
Microsoft System Tracks Taxis to ID Urban Transportation Problems
A Microsoft Research Asia team has demonstrated that tracking taxis via GPS could identify problems with a city's transportation network and help planners reduce traffic jams.
The researchers, led by project leader Yu Zheng, analyzed GPS data collected from 33,000 Beijing taxicabs in 2009 and 2010, looking for more than just areas of congestion. They also wanted to use the information to identify the underlying causes of traffic issues. To do this, they studied information about the way people traveled through the city, including the routes they used.
The researchers also analyzed the data to identify problems encountered in traveling among different regions of Beijing. For example, if many drivers take a roundabout route between two points, there might be a problem in the direct route.
In some cases, the research found that traffic problems aren't always in the busiest areas. For example, if many travelers must pass from one region through a second to get to a third, it might look like the middle area is busy and needs wider roads. However, Microsoft says, its study could indicate that a better solution is to create direct connections between the first and third regions.
The researchers tested their algorithms by comparing their findings over a period of time with changes to Beijing's transportation system. They learned that when the city made improvements to areas the study identified as having problems, traffic conditions got better. But when officials took action in other regions, conditions didn't improve.
Microsoft said its GPS-based tracking and traffic-analysis system could be useful in any city with a large number of taxicabs.
Great Leap Forward in Gesture-Control Technology
A US company has generated considerable buzz with a recently demonstrated motion controller planned for release this winter.
Leap Motion has developed the Leap gesture-control device, which plugs into a computer's USB port and comes with software for users to download. Leap turns the 8 cubic feet in front of it into an area in which users can control their computers with their gestures.
The technology tracks user motions in the control area with 200 times the precision of current gesture-control approaches, according to Leap Motion. The company says that Leap can track motions accurately to a hundredth of a millimeter and that users will be able to control its level of sensitivity.
During demonstrations of the system, Leap has translated user movements into corresponding motions on a video display with little or no latency.
According to Leap Motion, its system can follow as many items as are in its tracking zone and can even distinguish motions by each of a user's 10 fingers. This would enable very fine-grained control gestures such as using a pinching motion to zoom in on an image. The approach could also eliminate the need for users to make very expansive gestures, as some systems require.
Leap Motion says its goal is to make computers easier and more intuitive to work with. Potentially, it could be used in many ways, like the Kinect controller, which Microsoft originally developed for games but is now also utilized for other gesture-control functions.
A potential problem is that users might be tiring of motion-control devices, according to a survey by IGN Entertainment, an online media and services provider focused on gaming and entertainment.
Some industry observers say that the new technology is best suited for applications such as gaming and design but not as a replacement for the mouse in everyday computer use.
Leap Motion plans to begin selling its product late this year or in early 2013 for $69.99, and it will also make software developer kits and APIs available.
Hackers Inject Ads into Wikipedia Pages
Wikipedia is warning users that if they see a commercial advertisement on one of its pages, they might well have already been infected by malware.
The Wikipedia Foundation says victims may have downloaded a malicious, ad-injecting browser extension such as the "I want this" Google Chrome extension.
Security experts say this type of malware is associated with click fraud, in which an individual or application pretends to be a legitimate Internet user clicking on an advertisement to view it. Artificially producing many clicks can generate revenue for the network displaying the ad. Perpetrators could also use click fraud to deplete the budget of the company that has to pay the network showing its advertisement.
Wikipedia recommends that users who see ads on its pages run a scan with updated antivirus software and disable all browser add-ons to determine if they are causing the problem. However, this might not resolve the issue if the malware has other components on the victim's system that could infect the browser.
If these steps don't help, Wikipedia says, service providers may be injecting advertisements at the network level. In this case, the company noted, visiting affected websites via HTTP Secure sometimes eliminates the problematic ads.
The Wikipedia Foundation notes that it never runs advertisements except as part of its own annual fundraising efforts.
In Case the Fast Fourier Transform Isn't Fast Enough
MIT researchers have developed a data-processing algorithm designed to speed up the Fourier transform, a key mathematical approach used in fields such as computing, physics, and engineering.
The researchers, led by professors Dina Katabi and Piotr Indyk of MIT's Department of Electrical Engineering and Computer Science, have designed what they call the sparse Fourier transform (SFT).
The Fourier transform converts a time-dependent signal into a collection of sine and cosine waves that is frequency dependent. Users could then manipulate these waves for various purposes, such as removing unwanted noise from or compressing audio files.
However, the original algorithm ran slowly on computers. In response, researchers developed the fast Fourier transform in 1965. The FFT has been used in many areas, including MP3's compression of digital audio files.
The MIT researchers say their SFT can process data up to 100 times faster than the FFT primarily because it is used only on structured data like that found in many applications, including digital music. Such structured data contains a limited range of information, allowing the new algorithm to streamline its calculations.
The new approach would let systems process more information, produce results more quickly, and use less computing power. This makes the SFT particularly well suited for battery-powered mobile devices.