To leverage the pervasive smart mobile devices, a big challenge arises from the crowd's decentralized nature. How to effectively coordinate the individual members and make their functions into full play becomes a key issue. One essential aspect is to find reasonable coordination solutions for maximal group benefits without harming individual interests. We are taking new approaches in distributed AI and deep learning to address the challenge. Some ideas have been investigated for data-centric mobile crowd sensing applications.
With the high penetration rate of smart mobile devices, device-to-device (D2D), i.e., mobile-to-mobile, communications are becoming an appealing approach to support many proximity services, like social apps, location-based marketing, and public safety services. In particular, D2D can facilitate data dissemination aiming at delivering information a group of target users in a geographical region, such as in disaster alert, event notification, and classifieds distribution. Meanwhile, the popular social networks can provide valuable information, such as users' social characteristics and mobility patterns, to empower data dissemination. We developed social-aware energy-efficient D2D data dissemination algorithms, which can not only reduce energy cost and completion time, but also address the incentive constraint of participating users.
Nowadays, the mainstream mobile devices (smartphones, tablets, iPads, etc.) are equipped with multiple radio interfaces, like Wi-Fi, Bluetooth, and 3G/4G. Many users own multiple mobile devices. Have you ever thought how to make the best use of them when playing your favourite video game? There is a newly-minted Internet protocol, called multipath multipath transport control protocol (MPTCP), which can enable sending and receiving data across different network paths and interfaces at the same time. Apple's iOS 7 is the first large-scale use of this protocol. As MPTCP is designed for generic networks rather than the mobile network, especially, if you want to pool your mobile devices together, further extensions are needed to enhance the performance of this new protocol. We have proposed several enhancement mechanisms for MPTCP in a cooperative scenario that involves multiple multi-radio mobile devices. The performance of MPTCP with such enhancements significantly improves when compared with the original version.
In recent years, mobile network operators are facing overwhelming demands for high date rates and ever-emerging new applications. Meanwhile, cloud computing is achieving great success in empowering end users with rich experience by leveraging resource virtualization and sharing. The merging of cloud computing into the mobile domain creates the appealing paradigm of mobile cloud computing (MCC). You may be able to make money by leasing your redundant computational and networking resources. When will you be willing to share your resources? Get paid satisfactorily! But remember you are not a market "monopoly" but facing competitions from different providers. Thinking of the auctions at eBay, you must bid wisely to win. Resource sharing in MCC can be viewed as auctions too. You must run brilliant mechanisms to ensure that both MCC providers and end users are happy.