AI, digital banking, mobile commerce, and wearables were expected to be some of the hot topics this year at Mobile World Congress in Barcelona, Spain. As mobile demand increases with the implementation of 5G, developers are looking for more ways to make systems faster and cheaper, while also creating new and improved devices for mobile users. While the mobile world won’t gather this year in Barcelona, we’ll still likely see new product announcements in the weeks ahead. To keep you up-to-speed on the latest mobile advances, we flagged some of the cutting-edge research published recently in IEEE Xplore.
The Vehicular Cloud Provides Faster, Low-Cost Video Streaming
One way to handle the explosive growth in mobile traffic is with faster, low-cost video streaming through vehicular edge-caching, or what some people call the “vehicular cloud.”
In addition to increasing the number of caches in fixed locations or base stations, using public or private transportation as mobile relays and caches could help meet growing mobile demand, according to a research team from France. Advantages of edge caching include the ability to serve many requests locally, faster data transfer (less latency), and reduced load on the core network.
The researchers argue in a new study that mobile caches could be used for low-cost, delay-free video streaming. They performed trace-based simulations to support this contention, concluding that up to 60% of the original traffic could be offloaded by edge caching from the main infrastructure.
There are three key benefits to the vehicular cloud, according to Luigi Vigneri of the research team: it virtually extends the size of accessible local storage; it reduces capital and operational costs; and it opens the market to new network operators.
Some countries are already developing vehicular clouds, Vigneri said. In Portugal, a company called Veniam recently built the largest vehicular network in the world, which has increased the popularity of its public transportation system and the benefits that derive from that.
“They can offer Wi-Fi features in public transportation, increasing the number of passengers, reducing emissions, and generating additional revenue.”
Figure 1: A chart shows a sequence of three caches and the amount of data in end user buffer over time (shown in green). The region shaded in red shows when data is downloaded from the vehicle.
The more people access websites on their mobile devices, the more they may worry about the ability of those sites to track their location. A Missouri research team developed an app to address this problem called MoveWithMe. The proposed app generates human-like queries to hide the real users' locations and intentions when they are using location-based mobile services, helping increase their online privacy.
The MoveWithMe app works by generating decoys that behave like real humans. Each decoy has its own moving patterns, daily schedules, and social behaviors, which are different from the real user's.
Figure 2: An example of how MoveWithMe generates decoy users.
When the team tested the app for an entire day and compared it to the real user’s historical patterns, they found the app was able to mimic a real user’s schedule, such as lunch breaks and leaving for home at night. Yet the actual pattern was quite different from the real mobile user, effectively creating a decoy. The research team also found that data mining algorithms could only detect the decoy with 60% accuracy–only slightly higher than a random guess (50%). In comparison to randomly generated dummy profiles that can be detected with 95% accuracy, the MoveWithMe app was a significant improvement.
AR Headsets Aid Visually Impaired
More than 280 million people have moderate to severe vision impairment or blindness, and many rely on lens-based aids to improve their eyesight. A new app called Bright, which works on consumer augmented reality (AR) headsets like the Microsoft HoloLens, could help improve their vision more effectively and at a lower cost than traditional visual aids built with optical lenses, magnifying glasses, or magnified video feeds.
While traditional AR layers information on the user’s world view, this app uses a voice interface to process images. With a simple voice command (“read this”), users can instruct the app to read content from things like a book or TV screen, and the system will read any printed text out loud. The app can also use facial recognition to identify people, and users can pre-register familiar faces such as members of their family. In addition, an adjustable zoom magnifies the user’s field of view to help users see farther distances.
In the future, researchers hope to add text-to-speech and facial recognition. They are also working on a navigation-assistance system that would combine a depth-sensing camera with 360-degree auditory cues to help the user better understand his or her environment.
Figure 3: The Bright App Recognizes Familiar Faces.
Researchers have developed a mobile app to help farmers in India cut out the middlemen when they sell their products, increasing their profits and lowering prices for food purchasers.
Farming in India has been dominated in recent years by mediators in state-owned retail markets called “mandis,” who can take up to 70% of agricultural profits. Because of this system, farmers’ earnings have suffered and retailer costs have increased.
A research team from VIT Vellore in Tamil Nadu, India have proposed a mobile app called CropShop in hopes of helping India’s farming community by removing the need for mediators in agricultural transactions. The app will serve as a platform for growers to sell products to retailers or customers directly.
For more information on mobile advances, visit the IEEE Xplore digital library.
