Incentive Approaches for Data Dissemination in Autonomous Mobile Social Networks

• Main Author: Ning, Ting
• Language: EN
• Published: University of Louisiana at Lafayette 2014
• Subjects: Computer Science
• Online Access: http://pqdtopen.proquest.com/#viewpdf?dispub=3622949

<p> An autonomous mobile social network is formed by mobile users who share similar interests and connect with one another using the short range radios of their devices. Such networks not only re-assemble the real-world people interaction, but also can propagate data among mobile users. This dissertation centers on data dissemination in autonomous mobile social networks, where data fall into a range of interest types and each node may have one or multiple interests. The goal is to deliver data messages from source to nodes with corresponding interests.</p><p> Firstly, I have proposed an effective mechanism to track the value of a message under such a unique network setting with intermittent connectivity and multiple interest types. Given intermittent connections, credits are rewarded to the final deliverer only. Thus the value of a message for an intermediate node highly depends on its probability to deliver the message. Such probability itself is nontrivial to estimate. Moreover, a message is usually desired by multiple mobile users. Therefore, it can be potentially "sold" multiple times to different receivers. The effective interest contact probability captures the likelihood that a node contacts a sink of certain interest.</p><p> Secondly, I have proposed a novel incentive scheme for data dissemination in data pulling system based on a two-person bargaining game model to encourage cooperation among selfish nodes. The bargain process is formulated as a two-person cooperative game. Optimal Nash Solution is proposed for the bargain process and a greedy algorithm is developed to resolve the game and find out an optimal solution.</p><p> Thirdly, I address the stimulation problem for data dissemination in data pushing system and introduce the key idea of "virtual checks" to eliminate the needs of accurate knowledge about whom and how many credits data providers should pay. Both data packets and signed virtual checks can be traded between mobile nodes. Effective mechanisms are proposed to define virtual rewards for data packets and virtual checks and formulated nodal interaction as a two-player cooperative game.</p><p> Finally, extensive simulations have validated the viability of proposed incentive schemes based on real-world traces in terms of data delivery rate, delay and overhead.</p>