Research On Password Strength Evaluation Technology

Without cybersecurity,there is no national security.With the development of science and technology in recent years,more and more identity authentication technologies have been proposed.However,text passwords are still the most widely used authentication method due to their strong availability and strong variability.When setting a password,people will set different passwords according to their actual situation.With the advancement of technology,deep learning has been introduced into people's vision.It allows machines to do things that humans can't.In the current research on passwords,most of the researches on passwords simply design a password security protocol without paying attention to personal factors,and traditional researches on passwords are rarely combined with current popular deep learning algorithms.Given the above-mentioned problems,this article selects some subject databases with personal specific factors and researches the password strength evaluation technology from the theoretical perspective based on statistical theory and deep learning theory.This work is supported by the National Natural Science Foundation of China(NO.61902396,61802404,61802394),the Strategic Priority Research Program of Chinese Academy of Sciences(NO.XDC02040100),the Key Laboratory of Network Assessment Technology at Chinese Academy of Sciences and Beijing Key Laboratory of Network Security and Protection Technology.The main tasks are completed:(1): Collection,analysis,and processing of data.The 33 million pieces of password information collected in this article come from 7 databases leaked online.Passwords are generated by people and have a direct relationship with their behaviors.A password may be a strong password in one system,but a weak password in another system.This paper analyzes the relative strength of password strength,commonly used password analysis,and subject-based password exploration analysis on the collected databases.(2): Based on the traditional password guessing algorithm,a new password guessing model based on PCFG—T-PCFG.According to the traditional password,guessing algorithm does not pay attention to personal factors,this paper combines the PCFG algorithm to propose a topic-based password guessing model.It focuses on individuals setting their passwords based on topics(personal interests,religious beliefs,cultural backgrounds,etc.).(3): Based on a deep learning algorithm,a rule-based password guessing model,At PGAN,is proposed.With the rise of deep learning in recent years,this paper proposes a generative adversarial neural network model combining attention mechanisms,combining deep learning with password guessing.Based on the theoretical results of T-PCFG,we will further explore some rules of a personal password setting.The paper verifies the proposed algorithm experimentally at the end of the corresponding chapter.The model proposed in this paper is evaluated from the aspects of password guessability,model validity,and feasibility.Finally,the paper summarizes the research results and looks forward to furthering research based on existing problems.Figure 29,table 14,61 references.