Design And Implementation Of TOR Node Data Management System

In recent years,as Internet users have increasingly strict requirements for communication security,anonymous communication tools have been increasingly widely used.While protecting users'communication identity,their security and transmission efficiency have also become important indicators to measure anonymous tools.This thesis researches on the anonymous communication system,especially Tor.In the Tor network,transmission process is transparent to users,the node efficiency and security of the transmission link cannot be known by users,and low bandwidth and unstable nodes will affect the performance and security of Tor.At present,there is no mature platform for collecting and displaying the information of Tor nodes in the market.This thesis focuses on the operation log of Tor nodes,analyzes the characteristics and correlation of all kinds of information,and provides users with complete information of Tor network nodes by reorganizing and analyzing the data,so as to realize the Tor node data management system.The main contributions in this thesis are as follows:Firstly,this thesis studies the communication security and efficiency problems brought by the transparency of Tor network nodes under the background of anonymous communication,and proposes the requirements for integration of Tor node information based on the current research status.Then,the data sources are analyzed,the internal relationship of files is organized,and the overall architecture of the Tor node data management system is designed according to the requirements in this thesis.Secondly,this thesis presents an efficient data collection framework for parallel download and decompression of compressed files according to the file organization form and classification characteristics of the history log,and realizes complete data collection combined with the timing task processing,which provides support for data preprocessing and analysis.Thirdly,aiming at the history log,this thesis research and design of the data preprocessing and data analysis model through analyzing the advantages of processing massive data by big data technology.In this process,first,the raw log is preprocessed to be structured data,and the log information is reorganized vertically released by time.Then,the dimension of different organizational forms of file is unified to solve the data conflict,and data fusion is implemented with associated attributes.In addition,the complete history information of nodes is summarized transversely from the perspective of node,and the effective file retrieval strategy is designed according to the fingerprint characteristics to locate node information rapidly.At last,according to the real-time data file,the feature information of all dimensions of active real-time nodes on the Tor network is integrated,and the priority list of real-time nodes is established according to the consensus weight.Finally,this thesis designs and implements the Tor node data management platform based on the SSM framework.This system integrates and accesses the complete functions of timing data collection,data file browsing,historical log preprocessing and analysis,and analysis result display.And the system is tested from function and performance to prove its stability,safety and reliability.