Research On Mobile Data Offloading Through Opportunistic Mobile Networks

With the explosive growth of mobile data,a large number of users continue to request content from content service providers through cellular networks.However,due to the limited bandwidth of the cellular network,it may face problems such as traffic overload and congestion in the near future.In order to meet the increasing data needs of users,content service providers urgently need to take a series of measures to reduce traffic load.In order to solve the above problems,relevant researchers have proposed to use various high-capacity,low-cost complementary networks to offload data,such as Wi Fi networks and small cell networks.However,they all depend on infrastructure,and have the defects of limited coverage and high installation cost,which limits the use and promotion.The use of opportunistic mobile networks for data offloading is a very promising technology to alleviate traffic overload of the cellular nerwork.It uses user mobility to assist in offloading data,which is not restricted by infrastructure and has a wider range of application scenarios.Its data offloading efficiency depends to a great extent on the selection of seed nodes.There have been some studies on the selection of seed nodes,but they have not considered the freshness of the content.In fact,users are very sensitive to the freshness of the content.Inspired by this problem,freshness-aware seed selection problem is proposed.In recent years,social network analysis technology has provided a new research idea for data offloading.Some studies have used communities to improve the performance of data offloading.However,the current studies are limited to use static networks to simulate the social characteristics of nodes,which does not conform to the dynamic characteristics of real networks.Different from the existing studies,this dissertation proposes a new seed selection algorithm based on the temporal community from the temporal perspective,which ensures that all nodes can receive data while minimizing the mobile data traffic of the cellular network.This dissertation studies mobile data offloading technology through Opportunistic Mobile Networks.The main research work includes:(1)Considering the freshness of the content,a freshness-aware seed selection problem is proposed,and the optimal initial number of seeds and the maximum overall utility of the content are obtained.The freshness-aware seed selection problem is modeled as a utility optimization problem,which not only considers the transmission cost from the cellular network to the initial seeds,but also ensures that all nodes can get the content within thedeadline.Based on the analysis of the optimal strategy,two types of seed selection algorithms are proposed: greedy seed selection algorithm and decay-based seed selection algorithm.Finally,through extensive simulations in real data sets,the superiority of the seed selection algorithm is proved;(2)The seed selection problem based on the temporal community is studied.The problem is modeled as an optimization problem that minimizes the amount of content transmitted through the cellular network while ensuring that nodes can obtain content within the deadline.In order to solve this optimization problem,a greedy seed selection algorithm based on the centrality metric is first proposed,and then a community-based seed selection algorithm is proposed.Since the performance of the previous two algorithms is not good and does not consider the temporal characteristics of the community,this dissertation proposes a seed selection algorithm based on the temporal community,which uses hierarchical clustering to perform clustering.Through extensive simulations in real data sets,the superiority of the proposed seed selection algorithm based on the temporal community is proved.