Efficient and Quality-Aware Data Access in Mobile Opportunistic Networks

Open Access
Zhang, Xiaomei
Graduate Program:
Computer Science and Engineering
Doctor of Philosophy
Document Type:
Date of Defense:
May 13, 2016
Committee Members:
  • Guohong Cao, Dissertation Advisor
  • Guohong Cao, Committee Chair
  • George Kesidis, Committee Member
  • Sencun Zhu, Committee Member
  • Zhenhui Li, Outside Member
  • Mobile Opportunistic Networks
  • Data Transmission
  • Transient Community
  • Truth Analysis
  • Mobile Crowdsourcing
  • Task Allocation
  • Social Networks
Recent advances in hand-held mobile devices have sparked the emergence of mobile opportunistic networks. These networks enable direct communication between mobile devices by utilizing the short-range communication interfaces embedded on mobile devices, such as Bluetooth or Wi-Fi Direct. Since mobile opportunistic networks rely on peer-to-peer connections between mobile devices, the data access does not require support from network infrastructures (e.g., cellular towers or access points). Therefore, mobile opportunistic networks have been widely used to facilitate data access in circumstances where network infrastructures are not available, such as during natural disasters or military attacks. The specific goal of this dissertation is to develop comprehensive solutions for providing efficient and quality-aware data access in mobile opportunistic networks. In particular, two research problems are addressed in this dissertation, i.e., realizing efficient data transmission between nodes and improving the quality of data received by those nodes. Due to the dynamic network topologies and the unexpected failure of mobile devices, data transmission in mobile opportunistic networks is extremely difficult. To address this problem, this dissertation explores the social behavior patterns of mobile users and proposes social-aware solutions to achieve efficient data transmission. In addition, the data received by mobile devices may suffer from quality-awareness problem. For example, the data may be of bad quality because of the existence of unreliable mobile users, who provide low-quality data due to their limited capability to provide data, or some other malicious or subjective factors. Solutions for addressing this quality-awareness problem are also provided in this dissertation. First, we propose a data transmission strategy by exploiting and utilizing a social structure called transient community, i.e., the community with associated temporal information. Most existing community detection methods fail to detect transient communities due to their aggregation of contact information into a single weighted or unweighted network. Therefore, we propose a novel clustering method to detect transient communities by exploiting the pairwise contact processes between mobile users. Then, we propose a transient community-based data transmission strategy by utilizing the detected transient communities as the data transmission unit. Second, we design data transmission strategies by considering the diverse connectivity characteristics in mobile opportunistic networks. Previous works fail to consider diverse connectivity characteristics, since they neglect the ubiquitous existence of Transient Connected Components (TCCs), where nodes inside a TCC can reach each other by multi-hop wireless communications. Exploiting the special characteristics of TCCs can increase contact opportunities and significantly improve data transmission performance. Third, to improve the quality of data received by the destination, we provide solutions to adaptively collect data from mobile users through mobile opportunistic networks, especially when the existing data are not enough to ensure data quality or credibility. Considering the requirement on data credibility and the constraint of network resources, we propose two optimization problems to quantify the tradeoff between the enhanced data credibility and the increased network overhead. We then design resource-aware approaches for the proposed problems. Finally, we design an approach to improve data quality by considering the expertise of the mobile users, which is based on the fact that a user may only have expertise on some problems in some domains, but not others. The proposed expertise-aware approach relies on an optimization problem to maximize the probability that data are requested from users with the right expertise, while ensuring the work load does not exceed the processing capability of each user. A maximum likelihood estimation (MLE)-based method is further proposed to estimate the truth of data, so that data quality can be improved.