Design And Implementation Of User Behaviour Simulation System Based On Human-computer Interaction

With the popularization of computer technology and the increasing importance of cyberspace security,the demand for technical drills and tests of countries,companies,and organizations is increasing,too.Network testbed technology is the key technology for meeting the needs above.In network testbeds,the simulated user behavior is an important component of the authenticity of the virtual network environment.In recent years,new types of network test beds have been demanding for more diverse user behavior simulation.Traditional methods based on message playback and manual scripts are bad at meeting the needs of scalability,fidelity,and automation.In addition,the industry also lacks a method for assessing the fidelity of the simulated behavior.The user behavior simulation system designed and implemented in this paper will help solve these two problems.First of all,in order to clarify the needs of the user behavior simulation system,this paper analyzes the system's functional requirements,non-functional requirements,and overall business processes.Then the roles of the system and the processes and relationships of system functions are explained.Second,this paper studies the representation method for real user behavior.In order to overcome the shortcomings of traditional message playback methods at readability,configurability,and robustness,this paper analyzes the human-computer interaction process,and designs a behavioral trajectory representation method based on the state-operation chain and corresponding data acquisition software.Traditional scripting methods,which involve manual design for operation rules and finite state machines,are cumbersome and with low automation and poor scalability.To solve these problems,this paper designs a state matching method and an operation matching method based on the behavior state graph model,which achieves the automatically generation of the behavioral state graph based on user behavior trajectory.The graph describes the operation and state transition rules of the desktop application.Third,in order to evaluate the fidelity of the simulated behavior,this paper studies the identity authentication and recognition model based on mouse and keyboard operation,and evaluates the degree of fidelity of behavior simulation through identity homology of operating behavior.In the research of identity authentication and recognition model,this paper designs a model based on the residual network and micro-session decision fusion framework,and compares its performance with the traditional machine learning models'.Experimental results show that the performance of the designed model exceeds the traditional models under most tasks.Fourth,aiming at behavior simulation problems in different scenarios,this paper designs a variety of session behavior generation methods and two types of mouse and keyboard operation simulation methods,and adopts a framework of session-driven operation to effectively use behavior data and models.In order to meet the requirements of network testbeds for behavior recurrence,customization,control,and simulation,this paper designs a variety of methods for generating conversational behavior.In order to achieve realistic simulation of keyboard and mouse operation behavior,this paper proposes a sampling-based keystroke-type operation simulation method and a template-based mouse move operation simulation method,and carries out a fidelity evaluation test of the mouse movement operation simulation method.The results show the method achieves good fidelity.Finally,this paper designs and implements the user behavior simulation prototype system based on the above techniques,and performs performance testing on the key data processing flow of the system.The functions of each module of the system work well,and the background collection performance of the data acquisition program is good.The data processing speed of the user behavior modeling module is fast enough.In general,the system can meet the functional requirements and processing performance requirements in practical use.