Research On Data Delivery And Resource Allocation Mechanism In Mobile Social Networks

To satisfy the increasing user demands of communication and social interactions,the knowledge of social network has been comprehensively applied in the researches of mobile communication networks,which results in the emergence of Mobile Social Networks(MSNs).MSN is a mobile communication system consisting of the social relationships among mobile users.The people,things and environment in MSNs are connected with each other through the communication network and the social network.In MSNs,numerous researches design efficient communication and network protocols with social analysis methods to improve the performance of network and the quality of services(QoS).This paper focuses on the data delivery and resource allocation mechanism design.The researches of data delivery can provide mobile users with efficient and low cost information exchange and information diffusion.In addition,efficient resource allocation methods can optimize the management and the usage of network resources,such as communication,energy,storage,computation,etc.Consequently,it is important to improve the performance of data delivery in terms of delivery ratio and delay as well as the resource cost of communication,energy,cache and bandwidth.The following parts of this paper will discuss from the perspectives of the social network and communication network with the integrated optimization of data delivery and resource allocation: 1)A user clustering based information exchange mechanism;2)An energy saving information exchange mechanism;3)An energy charging and computing enabled information diffusion mechanism;4)A mobile edge caching enabled information diffusion mechanism.Information exchange is a unicast data delivery method,in which the user clustering characteristic is exploited to improve data delivery performance.However,the existing user clustering based Information exchange schemes are not scalable,which performance will degrade rapidly with the increase of the network scale.Based on the social characteristic of user clustering,this paper introduces a two-layer information exchange model,based on which the information will be forwarded in the macro layer or the micro layer to dynamically adaptted to the distance between the source user and the destination user.Then,a two-layer QoS-aware information exchange algorithm(TLD)is proposed to satisfy the QoS in terms of information exchange ratio,delay and data size,which improves the performance of information exchange and degrades the algorithm complexity.Since the information exchange in MSNs is based on the cooperative communication among mobile users,the information relay during the cooperative communication will induce the energy cost for mobile devices,which tremendously degrades the performance and the lifetime of the network.This paper further studies the energy saving problem of information exchange and proposes an energy saving information exchange scheme(ESD),aiming to minimize the total energy cost during information exchange.We attempt to reduce the energy cost in three phases,namely,the predict phase,the routing phase,and the retransmission phase.First,in the predict phase,we employ the time-vary characteristic of the destination user's clustering in the gather point to optimize the information exchange efficiency,which decreases the probability of retransmission.Second,in the routing phase,a routing scheme with the minimal energy cost is proposed.Third,in the retransmission phase,the number of retransmission is optimized,and the total energy cost of information exchange is minimized by integrating the above three phases.Although the energy saving scheme can reduce the energy cost of mobile devices,the limited battery of mobile devices still limits the performance of data delivery,especially in the information diffusion which requires considerable number of users to participate in the cooperative communication.This paper studies the information diffusion problem with the energy and computing resource allocation.We explore the collaboration of communication,computation and energy to improve the information diffusion ratio.This part consists of three works: 1)Based on the social knowledge of user clustering and community similarity,a relay selection algorithm is proposed to maximize the total capacity of information diffusion among the relay set.2)Considering the cooperation of algorithm computing and information diffusion,an information diffusion scheme is proposed to enhance the information diffusion efficiency by employing edge computing among the relay nodes.3)Considering the cooperation of energy harvesting and information diffusion,we propose the energy allocation schemes respectively for the base station and the relay nodes.On the side of the relay nodes,an energy allocation scheme is proposed to balance the energy consumption of data broadcasting and algorithm computing.On the side of base station,we integrate Radio Frequency based Energy Harvesting and beamforming technologies into MSNs.A Markov decision process based energy transmission scheme is designed to supplement devices' energy capacity,which increases the information diffusion ratio.Finally,the performance of information diffusion has been improved with the combination of the above three works.The bandwidth cost of information diffusion can be reduced by storing the popular information at the edge of network,but existing researches lack the analysis model for edge caching based information diffusion,and neglect the social characteristic of mobile users for the design of edge caching scheme.This paper proposes a Markov-chain-based model from the viewpoint of social network to analyze the impact of edge caching among edge nodes,i.e.,remote radio head(RRH)and user equipment(UE),on the performance of information diffusion,as well as the bandwidth consumption of radio access networks(RANs)and fronthaul.Then,we analyze the contact frequency between mobile users,and propose edge caching scheme among UEs to minimize the bandwidth consumption in the RANs during the information diffusion.Considering the information sharing among mobile users and the user clustering in the gather points,we further propose edge caching strategies among RRHs in terms of caching locations and times to minimize the bandwidth consumption of fronthaul and storage costs induced by the edge caching.This paper systematically studies the mechanism design of data delivery and resource allocation in MSNs.The research of data delivery consists of information exchange and information diffusion mechanisms,while the resource allocation mechanism focuses on the resource allocation of network resources in terms of communication,energy,computing,and storage.Based on the social network analysis methods,the proposed mechanism improves the performance of data delivery and reduces the network resource overhead such as communication,energy,bandwidth,and storage with the joint optimization of data delivery and resource allocation.