Capacity Analysis Of Social Relationships-Based Inhomogeneous Wireless Networks

Due to the advancement of communication technology,wireless ad hoc networks(WANETs),as an important part of wireless communications,keep sustainable development,and the related construction of infrastructures also keeps continuous improvement.Based on the fact that the distribution of nodes in the real-world presents a certain inhomogeneous feature,and social relationships affect the communication probability and content transmission of nodes,the capacity of inhomogeneous wireless networks with social relationships is a fundamental and meaningful problem.According to the order from basic theory to the forefront trend,this thesis analyzes the capacity of different inhomogeneous wireless networks scenarios,and based on which this thesis further analyzes the performance of secrecy capacity and delay,and describes the impact of social relationships on wireless communications.The basic theory refers to the analysis of network performance in the traditional host-to-host communication paradigm.Due to the fact that the access and transmission of contents have become the main part of Internet traffic,the forefront trend refers to the analysis of network performance in the host-to-content communication paradigm.This thesis designs capacity achieving schemes in different scenarios of inhomogeneous wireless networks,and discusses the impacts of related system parameters on these three metrics.The research content in this thesis is summarized as follows:First,this thesis studies the impacts of social relationships and inhomogeneous node distribution on the capacity of WANETs.In inhomogeneous WANETs,nodes follow a cluster topology,and the social relationships among nodes are modeled by a rank-based model,which makes nodes more likely to communicate with neighboring nodes.This thesis first calculates the average traffic distance from the source node to the destination node according to the rank-based model,and proposes a scheme to achieve the lower bound of the throughput capacity.Then a simplified community detection algorithm is proposed,and node communication is divided into two parts according to the distance between nodes: the quasi-strong ties and quasi-weak ties.Communications of quasi-strong ties represent the transmission between nodes in the same community(or social group),while communications of quasi-weak ties represent the transmission between nodes from different communities.Finally this thesis calculates the capacity of quasi-strong ties and that of quasi-weak ties,and reveals the relationship among network capacity,social relationships,and inhomogeneous node distribution.Secondly,based on the analysis of the first part and from the perspective of physicallayer security,this thesis studies the secrecy capacity of WANETs with social relationships.Similarly,the social relationships among nodes are modeled by a rank-based model,which makes legitimate nodes more likely to communicate with neighboring legitimate nodes.This thesis considers two distribution of legitimate nodes: cluster topology and homogeneous Poisson point distribution(PPP),and proposes transmission schemes for two distribution of legitimate nodes,respectively.This thesis considers two modes for passive eavesdroppers: noncolluding case and colluding case,and adopts the technique of self-interference cancelation to suppress the SINR(Signal-toInterference-plus-Noise Ratio)of eavesdroppers.Based on the distribution of legitimate nodes and modes of passive eavesdroppers,this thesis studies the secrecy capacity of four scenarios: the scenario of inhomogeneous legitimate nodes and noncolluding eavesdroppers,the scenario of inhomogeneous legitimate nodes and colluding eavesdroppers,the scenario of homogeneous legitimate nodes and noncolluding eavesdroppers,and the scenario of homogeneous legitimate nodes and colluding eavesdroppers.The theoretical results demonstrate the secrecy performance of the proposed transmission schemes,and reveal the relationship among secrecy capacity,social relationships,and legitimate node distribution.Finally,with the communication paradigm development of the host-to-content interaction,in information-centric networks(ICNs)with infrastructure support and social relationships,this thesis first designs a content delivery mechanism for host-tocontent interaction,and divides the network topology into the cluster-dense case and cluster-sparse case according to the relationship between the number of clusters and that of base stations.Then this thesis proposes a cell tessellation multihop transmission scheme,and derives the capacity and delay performance of ICNs in the clusterdense case and cluster-sparse case,respectively.This thesis further proves the tradeoff between capacity and delay.Then based on the Zipf's Law for content popularity distribution,we adopt Lagrange multiplier method to optimize the caching strategy for capacity and delay performance.The social relationships impact the distribution of content popularity,and the content popularity distribution reflects the strength of social relationships in the network to a certain extent.The theoretical results demonstrate the performance of the proposed multihop transmission scheme,and reveal the relationship among the performance of ICNs,inhomogeneous node distribution,content popularity distribution,and infrastructure support.