Fair Local Resource Management in Mobile Social Networks
MetadataShow full item record
As 5G made its global debut at the Winter Olympics in Pyeongchang, the world is one step closer to its mass commercial deployments. However, the practical application of 5G will be limited to dense urban areas due to economical or technological reasons. In rural or developing areas, universal mobile coverage and reliable connection will still be luxury. Even in the areas covered by mobile networks, the network infrastructure may become unreliable when emergency such as power outage or natural disasters strikes. Recently, mobile social networking (MSN) is emerging as a new paradigm to enable communication in such challenged environments by taking advantage of human mobility and social connection to provide intermittent connectivity among mobile users. Though research on MSN has been progressing for years, real implementations and adoptions of mobile social networks (MSNs) in the public are rarely seen today. As the reason for this situation is obviously complex due to that the MSN technology is still immature, this thesis takes the opportunity to explore possible reasons by conducting a comprehensive survey on the state-of-the-art works on MSN. The survey provides an analysis on the design requirements and presents a functional architecture of MSN systems. By reviewing the literature according to the building blocks of the architecture, the survey finally outlines several open issues that need to be addressed, which include network bootstrapping, real-life implementation, sensitivity study, user behavior and interaction, local resource management, and incentive provisioning. Among these open issues, local resource management is a nearly untouched fundamental problem in MSN. Due to user mobility and short transmission range, the available airtime for users to disseminate their data in a temporary group is typically limited in MSNs. The success of group formation and information dissemination in MSNs fully depends on users’ willingness and cooperation. Therefore, the limited airtime must be fairly allocated to the users in a group in order to resolve conflicts of interest. In addition, a mobile device normally has limited capacities in terms of e.g. energy, storage, processing and communication, compared to infrastructural devices. In consequence, local resource management in MSNs not only should consider the devices with data to send or receive, but also must not forget the helpers that contribute additionally in terms of local resources to help the others. The thesis first shows the possibility of implementing optimization objectives such as fairness on top of an candidate short-range technology (i.e. WiFi Direct). With this positivity in mind, the thesis focuses on designing fair resource allocation schemes for data dissemination among a group of mobile users using Nash bargaining theory. The design follows an incremental approach that allows the designed schemes to be applicable from two-node group to multiplenode group, from specific utility function to general utility function, from specific technology to general technologies, etc.