| In the era of cloud computing,users upload data to the cloud for storage or processing,or obtain the required data from the cloud.With the rapid growth of the number of network edge devices and the amount of edge data generated by them,the transmission and computing capabilities of the network have been improved.High requirements,and limited by the development of network hardware,cloud computing can no longer meet the requirements of future network users.Edge computing(Edge Computing,EC),as the next stage of cloud computing,sinks data and calculations from the cloud to edge servers near users,forming a three-tier cloud-edge-network architecture that will effectively solve the cloud model The network disadvantages such as insufficient bandwidth and insufficient real-time performance.Content Centric Network(Content Centric Network,CCN)is an important research direction of next-generation network architecture.It is different from the traditional TCP / IP network,decouples the binding relationship between network content and its network storage location,and realizes content-centric The distributed cache mechanism and named routing mechanism effectively reduce the network request delay and increase the user access rate.From the perspective of covering the entire network and ensuring the privacy of all nodes,this paper fully considers the node characteristics of different levels and different interaction relationships,and proposes a hierarchical heterogeneous trust model and a privacy protection strategy based on the trust model,so as to ensure the credibility and reliability of the entire network.This thesis accomplishes the following three areas:(1)Aiming at the architectural characteristics of edge computing and making full use of the interaction characteristics between nodes at different levels,a hierarchical heterogeneous generalized trust model is proposed.This trust model covers all levels and all interaction relationships,and calculates direct trust based on the evaluation indicators obtained by the interaction relationship.Direct trusts include service trusts,content trusts,and collaboration trusts.Between cross-layer nodes,the service success rate or content verification pass rate is used to calculate service trust or content trust,the forwarding pass rate is used to calculate the collaborative trust between nodes at the same layer,and indirect trust and synthetic trust are calculated through direct trust.Finally,whether the node is malicious is judged according to whether the comprehensive trust exceeds the threshold,so as to complete the identification of the malicious node.(2)Aiming at the content privacy protection of edge computing,an adaptive privacy protection strategy based on the trust model is proposed,and the privacy protection strategy formulates different keys according to the different levels of nodes and the different interaction relationships of nodes.For the cross-layer node interaction relationship,the key is calculated by considering the environment credibility,resource margin,and request content similarity.For the interaction of nodes at the same layer of the thing layer,the environment credibility and resource margin are considered to calculate the key.For same-layer node interactions at the edge layer,the environment trustworthiness is considered to calculate the key.Finally,nodes encrypt content through different keys and encryption algorithms,using asymmetric encryption across layers and symmetric encryption between the same layers,so as to complete privacy protection.(3)C++ simulation is applied to verify the trust model and privacy protection strategy proposed in this paper.The experimental results in the same topology show that the trust model proposed in this paper is better than the two-layer trust model(Two-way Trust Management System,TTMS)and its extension model in terms of performance parameters and application parameters,on the other hand,the cost of processing the same amount of data is much smaller than that of the comparison model.This proves that the trust model and privacy protection strategy proposed in this paper are superior. |
