Article(id=1327018220717978265, tenantId=1146029695717560320, journalId=1324759914146529478, issueId=1327017934683222337, articleNumber=null, orderNo=null, doi=10.7544/issn1000-1239.202550399, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=research-article, receivedDate=1748966400000, receivedDateStr=2025-06-04, revisedDate=1752595200000, revisedDateStr=2025-07-16, acceptedDate=null, acceptedDateStr=null, onlineDate=1794315398625, onlineDateStr=2026-11-10, pubDate=1759248000000, pubDateStr=2025-10-01, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1794315398625, onlineIssueDateStr=2026-11-10, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1794315398625, creator=13701087609, updateTime=1794315398625, updator=13701087609, issue=Issue{id=1327017934683222337, tenantId=1146029695717560320, journalId=1324759914146529478, year='2025', volume='62', issue='10', pageStart='2397', pageEnd='2650', issueExtLink='null', onlineDate='null', pubDate='1759248000000', pubDateStr='2025-10-01', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=null, createTime=1794315330429, creator='13701087609', updateTime=1794376951272, updator='13701087609', preIssue=null, nextIssue=null, articleTotal=null, ext={EN=IssueExt(id=1327276391986475384, tenantId=1146029695717560320, journalId=1324759914146529478, issueId=1327017934683222337, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1327276391986475385, tenantId=1146029695717560320, journalId=1324759914146529478, issueId=1327017934683222337, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null, downloadFileDto=null}, startPage=2495, endPage=2511, ext={EN=ArticleExt(id=1327018220978025114, articleId=1327018220717978265, tenantId=1146029695717560320, journalId=1324759914146529478, language=EN, title=Tri-Phase Correlation Verification Based Fabricated Link Attacks Detection in Low-Latency Agentic Networks, columnId=1327017938323878221, journalTitle=Journal of Computer Research and Development, columnName=Special Issue on Data Security and Privacy Protection in New Distributed Networks and Systems, runingTitle=null, highlight=null, articleAbstract=

The critical challenge of detecting fabricated link attacks in low-latency agentic network environments, where conventional detection methods face significant limitations is investigated in this paper. The fundamental difficulty stems from three key factors: 1) the measured performance differences between normal and fabricated links often exhibit marginal variations that fall below meaningful detection thresholds; 2) dynamic background traffic patterns introduce substantial interference that obscures genuine attack signatures; and 3) coordinated overload attacks deliberately distort measurement accuracy. These factors collectively degrade the detection performance of existing methods by significantly reducing the observable distinction between legitimate and malicious links. To address these challenges, we propose a novel three-phase detection framework based on correlated link verification. The first phase implements an effective delay transformation method that reduces measurement errors while converting raw measurements into meaningful link performance metrics. The second phase employs a multipath transmission simulation technique that amplifies subtle performance disparities between correlated links into statistically observable features. The final phase establishes robust detection thresholds through the combined application of extreme value theory and probability distribution fitting methods. Comprehensive simulation results validate that our proposed method consistently outperforms existing approaches across diverse network scales and attack scenarios. The framework demonstrates particular effectiveness in low-latency agentic environments, maintaining superior detection performance while meeting stringent timing requirements. Specifically, the framework achieves a 20% to 30% improvement in F1 score compared with state-of-the-art methods. Notably, even under ultra-low latency conditions (e.g., 0.1ms link delay), the proposed method maintains a high detection rate of 95.23% with a false positive rate as low as 0.28%. These advancements provide a practical solution for securing next-generation autonomous networks against sophisticated link fabrication attacks.

, authors=Feng Zhang, Hao Feng, Xiumin Zhou, Maochun Zhang, Zhenming Chen, Yuming Liu*, authorsList=Feng Zhang, Hao Feng, Xiumin Zhou, Maochun Zhang, Zhenming Chen, Yuming Liu, authorCompany=null, correspAuthors=Yuming Liu, authorNote=

born in 1977. PhD. His main research interests include network security and data science

born in 1978. PhD, associate professor. His main research interests include computing intelligence and distributed system

born in 2000. Master candidate. His main research interests include agentic network and AIoT

born in 2001. Master candidate. His main research interests include artificial intelligence and agentic network

born in 2000. Master candidate. His main research interests include artificial intelligence and agentic network

born in 1991. PhD. His main research interests include software-defined networking, programmable networks, and network security

, correspAuthorsNote=null, copyrightStatement=Copyright © 2025 Journal of Computer Research and Development. All rights reserved., copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=null, magXml=null, pdfUrl=null, pdf=null, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=null, mapNumber=null, fund=null), CN=ArticleExt(id=1327018226099270339, articleId=1327018220717978265, tenantId=1146029695717560320, journalId=1324759914146529478, language=CN, title=低时延智能体网络中基于三阶段关联验证的伪造链路攻击检测, columnId=1327017938470678863, journalTitle=计算机研究与发展, columnName=新型分布式网络与系统中的数据安全与隐私保护专题, runingTitle=null, highlight=null, articleAbstract=

面向低时延的智能体网络环境,研究了一种基于关联链路验证的伪造链路攻击检测方法. 在该网络环境下,正常链路与伪造链路在测量时延等关键指标上的统计特征差异较小,并且背景流量和协调器过载攻击的影响会放大测量误差,但缩小该差异会导致现有方法的检测性能骤降. 对此,提出一种基于关联链路验证的伪造链路攻击检测方法,该方法包含3个阶段:首先,采用一种有效时延转换方法缓解测量误差的影响,并将处理后的测量时延等转化为链路性能指标;然后,根据提出的多径传输模拟方法,将关联链路之间的性能差异转化为易于观测的统计特征;最后,基于极值理论和概率分布拟合方法确定统计特征的阈值,用于伪造链路的检测. 仿真实验结果表明,在不同的网络规模和攻击场景下,所提出的方法能够有效地检测低时延智能体网络中的伪造链路攻击,并且在检测性能方面明显优于现有的相关方法,F1分数提升达20%~30%. 特别地,当链路传输时延低至0.1 ms时,仍然能取得95.23%的检测率和0.28%的误判率,显示出良好的鲁棒性.

, authors=张峰, 俸皓, 周秀民, 章茂淳, 陈镇铭, 刘玉明*, authorsList=张峰, 俸皓, 周秀民, 章茂淳, 陈镇铭, 刘玉明, authorCompany=null, correspAuthors=刘玉明, authorNote=

1977年生. 博士. 主要研究方向为网络安全、数据科学

1978年生. 博士,副教授. 主要研究方向为计算智能、分布式系统

2000年生. 硕士研究生. 主要研究方向为智能体网络、人工智能物联网

2001年生. 硕士研究生. 主要研究方向为人工智能、智能体网络

2000年生. 硕士研究生. 主要研究方向为人工智能、智能体网络

1991年生. 博士. 主要研究方向为软件定义网络、可编程网络、网络安全

, correspAuthorsNote=
刘玉明()
, copyrightStatement=版权所有 © 2025 计算机研究与发展编辑部, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=BVcavkEOzxMMOwXroDexSQ==, magXml=tyLeppwdb1kP4/QdeLV4GA==, pdfUrl=null, pdf=cybjDI54HWgqVUKnlLHLog==, pdfFileSize=1974840, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=KGcTZLR357csTzaluekTAw==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=rb4ei/OerycICp9U9NdH3g==, mapNumber=null, fund=null)}, authors=[Author(id=1327018984005808281, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=zhangfeng@guet.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1327018984068722843, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018984005808281, language=EN, stringName=Feng Zhang, firstName=Feng, middleName=null, lastName=Zhang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004, bio={"content":"

born in 1977. PhD. His main research interests include network security and data science

"}, bioImg=null, bioContent=

born in 1977. PhD. His main research interests include network security and data science

, aboutCorrespAuthor=null), CN=AuthorExt(id=1327018984148414620, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018984005808281, language=CN, stringName=张峰, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=桂林电子科技大学计算机与信息安全学院 广西桂林 541004, bio={"img":"QPZ8J4/RKKLmybxkwngHAQ==","content":"

1977年生. 博士. 主要研究方向为网络安全、数据科学

"}, bioImg=QPZ8J4/RKKLmybxkwngHAQ==, bioContent=

1977年生. 博士. 主要研究方向为网络安全、数据科学

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1327018983921922197, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, xref=null, ext=[AuthorCompanyExt(id=1327018983934505110, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004), AuthorCompanyExt(id=1327018983942893719, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=桂林电子科技大学计算机与信息安全学院 广西桂林 541004)])]), Author(id=1327018984207134878, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1327018984278438048, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018984207134878, language=EN, stringName=Hao Feng, firstName=Hao, middleName=null, lastName=Feng, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004, bio={"content":"

born in 1978. PhD, associate professor. His main research interests include computing intelligence and distributed system

"}, bioImg=null, bioContent=

born in 1978. PhD, associate professor. His main research interests include computing intelligence and distributed system

, aboutCorrespAuthor=null), CN=AuthorExt(id=1327018984442015905, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018984207134878, language=CN, stringName=俸皓, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=桂林电子科技大学计算机与信息安全学院 广西桂林 541004, bio={"img":"nolLLZSPlwth3t/lt3ihpQ==","content":"

1978年生. 博士,副教授. 主要研究方向为计算智能、分布式系统

"}, bioImg=nolLLZSPlwth3t/lt3ihpQ==, bioContent=

1978年生. 博士,副教授. 主要研究方向为计算智能、分布式系统

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1327018983921922197, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, xref=null, ext=[AuthorCompanyExt(id=1327018983934505110, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004), AuthorCompanyExt(id=1327018983942893719, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=桂林电子科技大学计算机与信息安全学院 广西桂林 541004)])]), Author(id=1327018984517513379, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1327018984597205157, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018984517513379, language=EN, stringName=Xiumin Zhou, firstName=Xiumin, middleName=null, lastName=Zhou, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004, bio={"content":"

born in 2000. Master candidate. His main research interests include agentic network and AIoT

"}, bioImg=null, bioContent=

born in 2000. Master candidate. His main research interests include agentic network and AIoT

, aboutCorrespAuthor=null), CN=AuthorExt(id=1327018984664314022, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018984517513379, language=CN, stringName=周秀民, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=桂林电子科技大学计算机与信息安全学院 广西桂林 541004, bio={"img":"VkVZyN6tP5P3iDaTB7Th5Q==","content":"

2000年生. 硕士研究生. 主要研究方向为智能体网络、人工智能物联网

"}, bioImg=VkVZyN6tP5P3iDaTB7Th5Q==, bioContent=

2000年生. 硕士研究生. 主要研究方向为智能体网络、人工智能物联网

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1327018983921922197, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, xref=null, ext=[AuthorCompanyExt(id=1327018983934505110, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004), AuthorCompanyExt(id=1327018983942893719, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=桂林电子科技大学计算机与信息安全学院 广西桂林 541004)])]), Author(id=1327018984752394408, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, orderNo=3, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1327018984857252010, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018984752394408, language=EN, stringName=Maochun Zhang, firstName=Maochun, middleName=null, lastName=Zhang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004, bio={"content":"

born in 2001. Master candidate. His main research interests include artificial intelligence and agentic network

"}, bioImg=null, bioContent=

born in 2001. Master candidate. His main research interests include artificial intelligence and agentic network

, aboutCorrespAuthor=null), CN=AuthorExt(id=1327018984928555179, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018984752394408, language=CN, stringName=章茂淳, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=桂林电子科技大学计算机与信息安全学院 广西桂林 541004, bio={"img":"igXoN2sl3+JHtOS8MRvcyQ==","content":"

2001年生. 硕士研究生. 主要研究方向为人工智能、智能体网络

"}, bioImg=igXoN2sl3+JHtOS8MRvcyQ==, bioContent=

2001年生. 硕士研究生. 主要研究方向为人工智能、智能体网络

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1327018983921922197, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, xref=null, ext=[AuthorCompanyExt(id=1327018983934505110, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004), AuthorCompanyExt(id=1327018983942893719, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=桂林电子科技大学计算机与信息安全学院 广西桂林 541004)])]), Author(id=1327018985062772909, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, orderNo=4, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1327018985352179887, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018985062772909, language=EN, stringName=Zhenming Chen, firstName=Zhenming, middleName=null, lastName=Chen, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004, bio={"content":"

born in 2000. Master candidate. His main research interests include artificial intelligence and agentic network

"}, bioImg=null, bioContent=

born in 2000. Master candidate. His main research interests include artificial intelligence and agentic network

, aboutCorrespAuthor=null), CN=AuthorExt(id=1327018985419288752, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018985062772909, language=CN, stringName=陈镇铭, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=桂林电子科技大学计算机与信息安全学院 广西桂林 541004, bio={"img":"zot74ZXaSynezrhN0vft2Q==","content":"

2000年生. 硕士研究生. 主要研究方向为人工智能、智能体网络

"}, bioImg=zot74ZXaSynezrhN0vft2Q==, bioContent=

2000年生. 硕士研究生. 主要研究方向为人工智能、智能体网络

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1327018983921922197, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, xref=null, ext=[AuthorCompanyExt(id=1327018983934505110, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004), AuthorCompanyExt(id=1327018983942893719, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=桂林电子科技大学计算机与信息安全学院 广西桂林 541004)])]), Author(id=1327018985494786226, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, orderNo=5, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=ymliu@guet.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1327018985645781172, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018985494786226, language=EN, stringName=Yuming Liu, firstName=Yuming, middleName=null, lastName=Liu, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004, bio={"content":"

born in 1991. PhD. His main research interests include software-defined networking, programmable networks, and network security

"}, bioImg=null, bioContent=

born in 1991. PhD. His main research interests include software-defined networking, programmable networks, and network security

, aboutCorrespAuthor=null), CN=AuthorExt(id=1327018985721278645, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018985494786226, language=CN, stringName=刘玉明, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=桂林电子科技大学计算机与信息安全学院 广西桂林 541004, bio={"img":"lKp8Wv0iiRUbZkApFKp6+A==","content":"

1991年生. 博士. 主要研究方向为软件定义网络、可编程网络、网络安全

"}, bioImg=lKp8Wv0iiRUbZkApFKp6+A==, bioContent=

1991年生. 博士. 主要研究方向为软件定义网络、可编程网络、网络安全

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1327018983921922197, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, xref=null, ext=[AuthorCompanyExt(id=1327018983934505110, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004), AuthorCompanyExt(id=1327018983942893719, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=桂林电子科技大学计算机与信息安全学院 广西桂林 541004)])])], keywords=[Keyword(id=1327018985918410934, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, orderNo=1, keyword=agentic network), Keyword(id=1327018986002297015, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, orderNo=2, keyword=low-latency), Keyword(id=1327018986065211576, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, orderNo=3, keyword=fabricated link attacks), Keyword(id=1327018986153291961, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, orderNo=4, keyword=correlated link verification), Keyword(id=1327018986228789434, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, orderNo=5, keyword=extreme value theory), Keyword(id=1327018986333647035, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, orderNo=1, keyword=智能体网络), Keyword(id=1327018986392367292, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, orderNo=2, keyword=低时延), Keyword(id=1327018986459476157, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, orderNo=3, keyword=伪造链路攻击), Keyword(id=1327018986530779326, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, orderNo=4, keyword=关联链路验证), Keyword(id=1327018986602082495, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, orderNo=5, keyword=极值理论)], refs=[Reference(id=1327018993229082856, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=1, rfOrder=0, authorNames=null, journalName=null, refType=null, unstructuredReference=Liu Yan, Guo Bin, Li Nuo, et al. CrowdTransfer: Enabling crowd knowledge transfer in AIoT community[J]. IEEE Communications Surveys & Tutorials, 2024, 27(2): 1191−1237, articleTitle=null, refAbstract=null), Reference(id=1327018993287803113, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=2, rfOrder=1, authorNames=null, journalName=null, refType=null, unstructuredReference=Liu Sicong, Guo Bin, Fang Cheng, et al. Enabling resource-efficient AIoT system with cross-level optimization: A survey[J]. IEEE Communications Surveys & Tutorials, 2023, 26(1): 389−427, articleTitle=null, refAbstract=null), Reference(id=1327018993371689194, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=3, rfOrder=2, authorNames=null, journalName=null, refType=null, unstructuredReference=Chen Weize, You Ziming, Li Ran, et al. Internet of agents: Weaving a Web of heterogeneous agents for collaborative intelligence[C]//Proc of the Int Conf on Learning Representations (ICLR). San Diego, CA: OpenReview. net, 2025: 1−32, articleTitle=null, refAbstract=null), Reference(id=1327018993447186667, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=4, rfOrder=3, authorNames=null, journalName=null, refType=null, unstructuredReference=Agent Network Protocol Contributors. Agent Network Protocol (ANP). [EB/OL]. [2025-7-16]. https://github.com/Agent-network-protocol/AgentNetworkProtocol, articleTitle=null, refAbstract=null), Reference(id=1327018993510101228, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=5, rfOrder=4, authorNames=null, journalName=null, refType=null, unstructuredReference=Fu Bo, Smith W, Rizzo D, et al. Robust task scheduling for heterogeneous robot teams under capability uncertainty[J]. IEEE Transactions on Robotics, 2022, 39(2): 1087−1105, articleTitle=null, refAbstract=null), Reference(id=1327018993573015789, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=6, rfOrder=5, authorNames=null, journalName=null, refType=null, unstructuredReference=Soltani S, Amanlou A, Shojafar M, et al. Security of topology discovery service in SDN: Vulnerabilities and countermeasures[J]. IEEE Open Journal of the Communications Society, 2024, 5: 3410−3450, articleTitle=null, refAbstract=null), Reference(id=1327018993640124654, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=7, rfOrder=6, authorNames=null, journalName=null, refType=null, unstructuredReference=郑正, 徐明伟, 李琦, 等. 网络拓扑污染攻击防御机制研究[J]. 计算机研究与发展, 2018, 55(1): 207−216, articleTitle=null, refAbstract=null), Reference(id=1327018993707233519, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=7, rfOrder=7, authorNames=null, journalName=null, refType=null, unstructuredReference=Zheng Zheng, Xu Mingwei, Li Qi, et al. Defending against SDN network topology poisoning attacks[J]. Journal of Computer Research and Development, 2018, 55(1): 207−216 (in Chinese), articleTitle=null, refAbstract=null), Reference(id=1327018993778536688, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=8, rfOrder=8, authorNames=null, journalName=null, refType=null, unstructuredReference=Taheri M, Khorasani K, Shames I, et al. Undetectable cyber attacks on communication links in multi-agent cyber-physical systems[C]//Proc of the 59th IEEE Conf on Decision and Control (CDC). Piscataway, NJ: IEEE, 2020: 3764−3771, articleTitle=null, refAbstract=null), Reference(id=1327018993862422769, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=9, rfOrder=9, authorNames=null, journalName=null, refType=null, unstructuredReference=Ding Lifu, Yan Gangfeng. A survey of the security issues and defense mechanisms of multi-agent systems[J]. CAAI Transactions on Intelligent Systems, 2020, 15(3): 425−434, articleTitle=null, refAbstract=null), Reference(id=1327018993937920242, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=10, rfOrder=10, authorNames=null, journalName=null, refType=null, unstructuredReference=丁礼富, 严刚峰. 多智能体系统安全性问题及防御机制综述[J]. 信息与控制, 2018, 47(6): 651−660, articleTitle=null, refAbstract=null), Reference(id=1327018993996640499, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=10, rfOrder=11, authorNames=null, journalName=null, refType=null, unstructuredReference=Ding Lifu, Yan Gangfeng. A survey of security issues and defense mechanisms in multi-agent systems[J]. Information and Control, 2018, 47(6): 651−660 (in Chinses, articleTitle=null, refAbstract=null), Reference(id=1327018994072137972, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=11, rfOrder=12, authorNames=null, journalName=null, refType=null, unstructuredReference=Cao Yinzhi, Yu A F, Aday A, et al. Efficient repair of polluted machine learning systems via causal unlearning[C]//Proc of the 2018 on Asia Conf on Computer and Communications Security (ASIACSS). New York: ACM, 2018: 735−747, articleTitle=null, refAbstract=null), Reference(id=1327018994143441141, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=12, rfOrder=13, authorNames=null, journalName=null, refType=null, unstructuredReference=Zeng Zengri, Peng Wei, Zeng Detian. Improving the stability of intrusion detection with causal deep learning[J]. IEEE Transactions on Network and Service Management, 2022, 19(4): 4750−4763, articleTitle=null, refAbstract=null), Reference(id=1327018994223132918, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=13, rfOrder=14, authorNames=null, journalName=null, refType=null, unstructuredReference=Smyth D, McSweeney S, O’Shea D, et al. Detecting link fabrication attacks in software-defined networks[C]//Proc of the 26th Int Conf on Computer Communication and Networks (ICCCN). Piscataway, NJ: IEEE, 2017: 1(1): 1−8, articleTitle=null, refAbstract=null), Reference(id=1327018994302824695, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=14, rfOrder=15, authorNames=null, journalName=null, refType=null, unstructuredReference=Al Salti I, Zhang N. LINK-GUARD: An effective and scalable security framework for link discovery in SDN networks[J]. IEEE Access, 2022, 10(10): 130233−130252, articleTitle=null, refAbstract=null), Reference(id=1327018994361544952, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=15, rfOrder=16, authorNames=null, journalName=null, refType=null, unstructuredReference=Soltani S, Shojafar M, Mostafaei H, et al. Link latency attack in software-defined networks[C]//Proc of the 17th Int Conf on Network and Service Management (CNSM). Piscataway, NJ: IEEE, 2021, 1(1): 187−193, articleTitle=null, refAbstract=null), Reference(id=1327018994428653818, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=16, rfOrder=17, authorNames=null, journalName=null, refType=null, unstructuredReference=Soltani S, Shojafar M, Mostafaei H, et al. Real-time link verification in software-defined networks[J]. IEEE Transactions on Network and Service Management, 2023, 20(3): 3596−3611, articleTitle=null, refAbstract=null), Reference(id=1327018994491568379, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=17, rfOrder=18, authorNames=null, journalName=null, refType=null, unstructuredReference=Skowyra R, Xu Lei, Gu Guofei, et al. Effective topology tampering attacks and defenses in software-defined networks[C]//Proc of the 48th Annual IEEE/IFIP Int Conf on Dependable Systems and Networks (DSN). Piscataway, NJ: IEEE, 2018: 374−385, articleTitle=null, refAbstract=null), Reference(id=1327018994550288636, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=18, rfOrder=19, authorNames=null, journalName=null, refType=null, unstructuredReference=Marin E, Bucciol N, Conti M. An in-depth look into SDN topology discovery mechanisms: Novel attacks and practical countermeasures[C]//Proc of the 2019 ACM SIGSAC Conf on Computer and Communications Security, New York, ACM, 2019: 1101−1114, articleTitle=null, refAbstract=null), Reference(id=1327018994625786109, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=19, rfOrder=20, authorNames=null, journalName=null, refType=null, unstructuredReference=Chou L D, Liu C C, Lai M S, et al. Behavior anomaly detection in SDN control plane: A case study of topology discovery attacks[J]. Wireless Communications and Mobile Computing, 2020, 2020(1): 8898949, articleTitle=null, refAbstract=null), Reference(id=1327018994697089278, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=20, rfOrder=21, authorNames=null, journalName=null, refType=null, unstructuredReference=Huang Xinli, Shi Peng, Liu Yufei, et al. Towards trusted and efficient SDN topology discovery: A lightweight topology verification scheme[J]. Computer Networks, 2020, 170: 107119, articleTitle=null, refAbstract=null), Reference(id=1327018994768392447, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=21, rfOrder=22, authorNames=null, journalName=null, refType=null, unstructuredReference=Shrivastava P, Agarwal A, Kataoka K. Detection of topology poisoning by silent relay attacker in SDN[C]//Proc of the 24th annual Int Conf on Mobile Computing and Networking. New York: ACM, 2018: 792−794, articleTitle=null, refAbstract=null), Reference(id=1327018994822918400, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=22, rfOrder=23, authorNames=null, journalName=null, refType=null, unstructuredReference=Wang Jiadai, Liu Jiajia, Guo Hongzhi, et al. Deep reinforcement learning for securing software defined industrial networks with distributed control plane[J]. IEEE Transactions on Industrial Informatics, 2022, 18(6): 4275−4285, articleTitle=null, refAbstract=null), Reference(id=1327018994881638657, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=23, rfOrder=24, authorNames=null, journalName=null, refType=null, unstructuredReference=Kong Dezhang, Shen Yi, Chen Xiang, et al. Combination attacks and defenses on SDN topology discovery[J]. IEEE/ACM Transactions on Networking, 2023, 31(2): 904−919, articleTitle=null, refAbstract=null), Reference(id=1327018994936164610, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=24, rfOrder=25, authorNames=null, journalName=null, refType=null, unstructuredReference=Azzouni A, Boutaba R, Trang N T M, et al. sOFTDP: Secure and efficient OpenFlow topology discovery protocol[C]//Proc of 2018 IEEE/IFIP Network Operations and Management Symp. Piscataway, NJ: IEEE, 2018: 1−7, articleTitle=null, refAbstract=null), Reference(id=1327018995003273475, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=25, rfOrder=26, authorNames=null, journalName=null, refType=null, unstructuredReference=Nehra A, Tripathi M, Gaur M S, et al. SLDP: A secure and lightweight link discovery protocol for software defined networking[J]. Computer Networks, 2019, 150: 102−116, articleTitle=null, refAbstract=null), Reference(id=1327018995061993732, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=26, rfOrder=27, authorNames=null, journalName=null, refType=null, unstructuredReference=Jia Yongze, Xu Lei, Yang Yuwang, et al. Lightweight automatic discovery protocol for OpenFlow-based software defined networking[J]. IEEE Communications Letters, 2020, 24(2): 312−315, articleTitle=null, refAbstract=null), Reference(id=1327018995120713989, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=27, rfOrder=28, authorNames=null, journalName=null, refType=null, unstructuredReference=Gu Yongpu, Li Dong, Yu Junqing. IM-OFDP: An improved openflow-based topology discovery protocol for software defined network[C]//Proc of 2020 IFIP Networking Conf (Networking). Piscataway, NJ: IEEE, 2020: 628−630, articleTitle=null, refAbstract=null), Reference(id=1327018995179434246, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=28, rfOrder=29, authorNames=null, journalName=null, refType=null, unstructuredReference=Hauser F, Schmidt M, Haberle M, et al. P4-MACsec: Dynamic topology monitoring and data layer protection with MACsec in P4-based SDN[J]. IEEE Access, 2020, 8: 58845−58858, articleTitle=null, refAbstract=null), Reference(id=1327018995242348807, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=29, rfOrder=30, authorNames=null, journalName=null, refType=null, unstructuredReference=Jero S, Koch W, Skowyra R, et al. Identifier binding attacks and defenses in software-defined networks[C]//Proce of the 26th USENIX Conf on Security Symp. Berkeley, CA: USENIX Association, 2017: 415−432, articleTitle=null, refAbstract=null), Reference(id=1327018995322040584, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=30, rfOrder=31, authorNames=null, journalName=null, refType=null, unstructuredReference=Zhou Qizhao, Yu Junqing, Li Dong. A dynamic and lightweight framework to secure source addresses in the SDN-based networks[J]. Computer Networks, 2021, 193: 108075, articleTitle=null, refAbstract=null), Reference(id=1327018995389149449, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=31, rfOrder=32, authorNames=null, journalName=null, refType=null, unstructuredReference=Meena R C, Bhatia S, Jhaveri R H, et al. HyPASS: Design of hybrid-SDN prevention of attacks of source spoofing with host discovery and address validation[J]. Physical Communication, 2022, 55: 101902, articleTitle=null, refAbstract=null), Reference(id=1327018995460452619, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=32, rfOrder=33, authorNames=null, journalName=null, refType=null, unstructuredReference=Haan L, Ferreira A. Extreme Value Theory: An Introduction[M]. New York: Springer, 2006: 65−126, articleTitle=null, refAbstract=null), Reference(id=1327018995523367180, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=33, rfOrder=34, authorNames=null, journalName=null, refType=null, unstructuredReference=Siffer A, Fouque P A, Termier A, et al. Anomaly detection in streams with extreme value theory[C]//Proc of the 23rd ACM SIGKDD Int Conf on Knowledge Discovery and Data Mining (KDD). New York: ACM, 2017: 1067−1075, articleTitle=null, refAbstract=null), Reference(id=1327018995615641869, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=34, rfOrder=35, authorNames=null, journalName=null, refType=null, unstructuredReference=Mininet Project Contributors. An instant virtual network on your laptop (or other PC)[EB/OL]. [2025-07-16]. https://mininet.org/, articleTitle=null, refAbstract=null), Reference(id=1327018997322723598, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=35, rfOrder=36, authorNames=null, journalName=null, refType=null, unstructuredReference=Scapy community. Scapy: The Python-based interactive packet manipulation program & library[EB/OL]. [2025-7-16]. https://github.com/secdev/scapy, articleTitle=null, refAbstract=null), Reference(id=1327018997427581199, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=36, rfOrder=37, authorNames=null, journalName=null, refType=null, unstructuredReference=Arturo Filastò. PyPcap - Python libpcap module. [EB/OL]. [2025-07-16]. https://github.com /pynetwork/pypcap, articleTitle=null, refAbstract=null), Reference(id=1327018997490495760, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=37, rfOrder=38, authorNames=null, journalName=null, refType=null, unstructuredReference=ESnet/Lawrence Berkeley National Laboratory. iPerf-The ultimate speed test tool for TCP UDP and SCTP[EB/OL]. [2025-07-16]. https://iperf.fr, articleTitle=null, refAbstract=null)], funds=null, companyList=[AuthorCompany(id=1327018983921922197, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, xref=null, ext=[AuthorCompanyExt(id=1327018983934505110, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004), AuthorCompanyExt(id=1327018983942893719, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=桂林电子科技大学计算机与信息安全学院 广西桂林 541004)])], figs=[ArticleFig(id=1327018988414021824, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, label=Figure 1, caption=Illustration of topology discovery and fabricated link attacks in agentic networks, figureFileSmall=IM2Y/qoHmRZ88IPGqAMHWw==, figureFileBig=KGcTZLR357csTzaluekTAw==, tableContent=null), ArticleFig(id=1327018988514685121, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, label=图1, caption=智能体网络中拓扑发现与伪造链路攻击示意图, figureFileSmall=IM2Y/qoHmRZ88IPGqAMHWw==, figureFileBig=KGcTZLR357csTzaluekTAw==, tableContent=null), ArticleFig(id=1327018988619542722, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, label=Figure 2, caption=Illustration of correlated links, figureFileSmall=P20kJkwZKMdMHo9lW5X5GQ==, figureFileBig=4DbEaTSxojid2rhKNtG/bQ==, tableContent=null), ArticleFig(id=1327018988716011715, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, label=图2, caption=关联链路示意图, figureFileSmall=P20kJkwZKMdMHo9lW5X5GQ==, figureFileBig=4DbEaTSxojid2rhKNtG/bQ==, tableContent=null), ArticleFig(id=1327018988871200964, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, label=Figure 3, caption=Ilustration of fabricated link detection method based on correlated link verification, figureFileSmall=ryzQypKZx+nOI3xx1CdOfg==, figureFileBig=krEVBH2+kymdA2VCm3ZWUA==, tableContent=null), ArticleFig(id=1327018988946698437, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, label=图3, caption=基于CLV的伪造链路检测方法示意图, figureFileSmall=ryzQypKZx+nOI3xx1CdOfg==, figureFileBig=krEVBH2+kymdA2VCm3ZWUA==, tableContent=null), ArticleFig(id=1327018989009612998, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, label=Figure 4, caption=Impact of varying super-parameters on detection performance of CLV method, figureFileSmall=hTdv2w/mz+t71jGgGEj3/Q==, figureFileBig=QOtvLkLZaXuAchtFCCxD6A==, tableContent=null), ArticleFig(id=1327018989097693383, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, label=图4, caption=不同超参数对CLV方法检测性能的影响, figureFileSmall=hTdv2w/mz+t71jGgGEj3/Q==, figureFileBig=QOtvLkLZaXuAchtFCCxD6A==, tableContent=null), ArticleFig(id=1327018989173190856, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, label=Figure 5, caption=Experimental topology, figureFileSmall=eYO0levN6bqCaVrQm5Vh6w==, figureFileBig=B614H9uvhHauOdL/k3Vk4Q==, tableContent=null), ArticleFig(id=1327018989319991497, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, label=图5, caption=实验拓扑, figureFileSmall=eYO0levN6bqCaVrQm5Vh6w==, figureFileBig=B614H9uvhHauOdL/k3Vk4Q==, tableContent=null), ArticleFig(id=1327018989420654794, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, label=Figure 6, caption=Distribution of measured delay before and after introducing fabricated link in varying attack scenarios, figureFileSmall=l4ymJqPJ+WpqXZsT93jryw==, figureFileBig=PA8iM8/Q6hWs7XMocYxtVw==, tableContent=null), ArticleFig(id=1327018989491957963, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, label=图6, caption=不同攻击场景下引入伪造链路前后测量时延的分布, figureFileSmall=l4ymJqPJ+WpqXZsT93jryw==, figureFileBig=PA8iM8/Q6hWs7XMocYxtVw==, tableContent=null), ArticleFig(id=1327018989554872524, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, label=Figure 7, caption=Time series of measured link quality under varying attack scenarios, figureFileSmall=gXAV9jw8sn3S6UsTXGHGSQ==, figureFileBig=9HabNcugiTfE07rMOaEsFA==, tableContent=null), ArticleFig(id=1327018989621981389, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, label=图7, caption=不同攻击场景下测量的链路质量的时间序列, figureFileSmall=gXAV9jw8sn3S6UsTXGHGSQ==, figureFileBig=9HabNcugiTfE07rMOaEsFA==, tableContent=null), ArticleFig(id=1327018989689090254, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, label=Figure 8, caption=Time series of entropy values for trajectory sequences of correlated links under varying attack scenarios, figureFileSmall=GvIF6vhcUaEarMZ1JRpkPQ==, figureFileBig=74WGH5rhA0GFO8TECydoQg==, tableContent=null), ArticleFig(id=1327018989747810511, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, label=图8, caption=不同攻击场景下关联链路轨迹序列熵值的时间序列, figureFileSmall=GvIF6vhcUaEarMZ1JRpkPQ==, figureFileBig=74WGH5rhA0GFO8TECydoQg==, tableContent=null), ArticleFig(id=1327018989823307984, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, label=Figure 9, caption=Distribution of K values for trajectory sequences of correlated links under varying attack scenarios, figureFileSmall=Q+lDZj3BH9m8ygbvCfel3Q==, figureFileBig=1SBQHjFqMtgfQHT6p4xIYQ==, tableContent=null), ArticleFig(id=1327018989902999761, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, label=图9, caption=不同攻击场景下关联链路轨迹序列$K$值的分布, figureFileSmall=Q+lDZj3BH9m8ygbvCfel3Q==, figureFileBig=1SBQHjFqMtgfQHT6p4xIYQ==, tableContent=null), ArticleFig(id=1327018989978497234, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, label=Figure 10, caption=Comparison of detection performance with related methods across varying network scales, figureFileSmall=rCYEdZoN6FuEtOsc8YC/3A==, figureFileBig=6/kAU8sBj2t3ifdP9NyHNg==, tableContent=null), ArticleFig(id=1327018990041411795, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, label=图10, caption=不同网络规模下相关方法检测性能的对比, figureFileSmall=rCYEdZoN6FuEtOsc8YC/3A==, figureFileBig=6/kAU8sBj2t3ifdP9NyHNg==, tableContent=null), ArticleFig(id=1327018990133686484, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, label=Figure 11, caption=Detection performance of CLV-RAW method under varying fabricated link counts, figureFileSmall=oIlrCnyus6aGbgD0rxm+KQ==, figureFileBig=5LMMAfZSQq8PQhrxmEnCug==, tableContent=null), ArticleFig(id=1327018990192406741, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, label=图11, caption=不同伪造链路数量下CLV-RAW方法的检测性能, figureFileSmall=oIlrCnyus6aGbgD0rxm+KQ==, figureFileBig=5LMMAfZSQq8PQhrxmEnCug==, tableContent=null), ArticleFig(id=1327018990246932694, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, label=Figure 12, caption=Detection performance of CLV-OPT method under varying fabricated link counts, figureFileSmall=Yq6ZzcxhNYO4/ARq7JpaZw==, figureFileBig=BsSPFdH833+CzS+eB6qosA==, tableContent=null), ArticleFig(id=1327018990309847255, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, label=图12, caption=不同伪造链路数量下CLV-OPT方法的检测性能, figureFileSmall=Yq6ZzcxhNYO4/ARq7JpaZw==, figureFileBig=BsSPFdH833+CzS+eB6qosA==, tableContent=null), ArticleFig(id=1327018990372761816, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, label=Figure 13, caption=Detection performance of CLV-RAW method under varying link delay, figureFileSmall=vCuyp9sMsQb83gvXs2irkg==, figureFileBig=Z4LViGj/YTEeA7qx+SFrqw==, tableContent=null), ArticleFig(id=1327018990431482073, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, label=图13, caption=不同链路时延下CLV-RAW方法的检测性能, figureFileSmall=vCuyp9sMsQb83gvXs2irkg==, figureFileBig=Z4LViGj/YTEeA7qx+SFrqw==, tableContent=null), ArticleFig(id=1327018990486008026, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, label=Figure 14, caption=Detection performance of CLV-OPT method under varying link delay, figureFileSmall=QWg9LbsMymjV8DtZh8O4kw==, figureFileBig=Ku8IKBqDmWxkyIN0ZaCNYA==, tableContent=null), ArticleFig(id=1327018990540533979, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, label=图14, caption=不同链路时延下CLV-OPT方法的检测性能, figureFileSmall=QWg9LbsMymjV8DtZh8O4kw==, figureFileBig=Ku8IKBqDmWxkyIN0ZaCNYA==, tableContent=null), ArticleFig(id=1327018990595059932, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, label=Table 1, caption=Notations and Description, figureFileSmall=null, figureFileBig=null, tableContent=
符号含义
$T$关于时延的随机变量
$D$用于训练模型的数据集
$I$示性函数
$Z$决策函数,判断链路是否为伪造链路
$\delta $检测阈值
$d$链路时延
$l$网络中的一条链路
$w$链路质量的评分,取值为0~1
$S$由0和1组成的轨迹序列,表示多径传输模拟方法的选路结果
$H$轨迹序列中0和1分布的香农熵值
$K$轨迹序列中连续1的最大长度
), ArticleFig(id=1327018990653780189, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, label=表1, caption=

符号及描述

, figureFileSmall=null, figureFileBig=null, tableContent=
符号含义
$T$关于时延的随机变量
$D$用于训练模型的数据集
$I$示性函数
$Z$决策函数,判断链路是否为伪造链路
$\delta $检测阈值
$d$链路时延
$l$网络中的一条链路
$w$链路质量的评分,取值为0~1
$S$由0和1组成的轨迹序列,表示多径传输模拟方法的选路结果
$H$轨迹序列中0和1分布的香农熵值
$K$轨迹序列中连续1的最大长度
), ArticleFig(id=1327018990729277662, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, label=Table 2, caption=Parameter Settings in Varying Attack Scenarios, figureFileSmall=null, figureFileBig=null, tableContent=
攻击
场景
传输时
延/ms
中继
模块
转发
时延
链路是
否过载
背景流量协调器
过载攻击
Case 15PyPcap低否无无
Case 20.5Scapy高否无无
Case 30.5PyPcap低是35~45 Mbps无
Case 40.5PyPcap低否无无
Case 50.5PyPcap低否无1000报文/s
Case 60.5PyPcap低否20~30 Mbps1000报文/s
), ArticleFig(id=1327018990792192223, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, label=表2, caption=

不同攻击场景下的参数配置

, figureFileSmall=null, figureFileBig=null, tableContent=
攻击
场景
传输时
延/ms
中继
模块
转发
时延
链路是
否过载
背景流量协调器
过载攻击
Case 15PyPcap低否无无
Case 20.5Scapy高否无无
Case 30.5PyPcap低是35~45 Mbps无
Case 40.5PyPcap低否无无
Case 50.5PyPcap低否无1000报文/s
Case 60.5PyPcap低否20~30 Mbps1000报文/s
), ArticleFig(id=1327018990863495392, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, label=Table 3, caption=Description of the Experimental Datasets, figureFileSmall=null, figureFileBig=null, tableContent=
攻击场景Agent数量伪造链路数量采样周期数
Case 1~ 6511500
Case 4~ 6511500
Case 610, 20, …, 10016000
Case 6301, 3, 6, 96000
样本特征分类类别类别比例训练与测试比例
$ ({d}_{\rm cs},{d}_{\rm sc},{d}_{l},load) $正常链路,伪造链路3∶12∶1
), ArticleFig(id=1327018990964158689, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, label=表3, caption=

实验数据集描述

, figureFileSmall=null, figureFileBig=null, tableContent=
攻击场景Agent数量伪造链路数量采样周期数
Case 1~ 6511500
Case 4~ 6511500
Case 610, 20, …, 10016000
Case 6301, 3, 6, 96000
样本特征分类类别类别比例训练与测试比例
$ ({d}_{\rm cs},{d}_{\rm sc},{d}_{l},load) $正常链路,伪造链路3∶12∶1
), ArticleFig(id=1327018991035461858, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, label=Table 4, caption=Comparison of Detection Performance for Relative Methods in Case 4 Scenario, figureFileSmall=null, figureFileBig=null, tableContent=
检测
指标
TopoGuard+[17]TrustTopo[20]RLV[16]CLV
MLPRFKNNSVMRAWOPT
TPR4.0056.962.7258.7260.7258.1899.3099.00
FPR0.3012.217.3013.9816.9312.7600.40
Pr93.0282.3478.5481.3478.8382.56100100
F17.6767.3069.7568.2068.6068.2699.6599.60
), ArticleFig(id=1327018991115153635, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, label=表4, caption=

Case 4场景下相关方法的检测性能对比

, figureFileSmall=null, figureFileBig=null, tableContent=
检测
指标
TopoGuard+[17]TrustTopo[20]RLV[16]CLV
MLPRFKNNSVMRAWOPT
TPR4.0056.962.7258.7260.7258.1899.3099.00
FPR0.3012.217.3013.9816.9312.7600.40
Pr93.0282.3478.5481.3478.8382.56100100
F17.6767.3069.7568.2068.6068.2699.6599.60
), ArticleFig(id=1327018992801263844, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, label=Table 5, caption=Comparison of Detection Performance for Relative Methods in Case 5 Scenario, figureFileSmall=null, figureFileBig=null, tableContent=
检测
指标
TopoGuard+[17]TrustTopo[20]RLV[16]CLV
MLPRFKNNSVMRAWOPT
TPR7.8553.4571.0263.7164.5469.4991.7099.60
FPR4.9518.0525.8017.1318.8123.9500
Pr61.3373.9373.5479.4278.0875.80100100
F113.9262.0472.2670.7170.6772.1895.6599.80
), ArticleFig(id=1327018992914510053, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, label=表5, caption=

Case 5场景下相关方法的检测性能对比

, figureFileSmall=null, figureFileBig=null, tableContent=
检测
指标
TopoGuard+[17]TrustTopo[20]RLV[16]CLV
MLPRFKNNSVMRAWOPT
TPR7.8553.4571.0263.7164.5469.4991.7099.60
FPR4.9518.0525.8017.1318.8123.9500
Pr61.3373.9373.5479.4278.0875.80100100
F113.9262.0472.2670.7170.6772.1895.6599.80
), ArticleFig(id=1327018992985813222, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=EN, label=Table 6, caption=Comparison of Detection Performance for Relative Methods in Case 6 Scenario, figureFileSmall=null, figureFileBig=null, tableContent=
检测
指标
TopoGuard+[17]TrustTopo[20]RLV[16]CLV
MLPRFKNNSVMRAWOPT
TPR8.9554.0064.8464.7462.9668.3098.7099.10
FPR5.7520.1525.3325.6025.4028.7300
Pr60.8872.8371.9672.0771.6770.81100100
F115.61620268.2168.2167.0369.5399.3599.55
), ArticleFig(id=1327018993065504999, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, language=CN, label=表6, caption=

Case 6场景下相关方法的检测性能对比

, figureFileSmall=null, figureFileBig=null, tableContent=
检测
指标
TopoGuard+[17]TrustTopo[20]RLV[16]CLV
MLPRFKNNSVMRAWOPT
TPR8.9554.0064.8464.7462.9668.3098.7099.10
FPR5.7520.1525.3325.6025.4028.7300
Pr60.8872.8371.9672.0771.6770.81100100
F115.61620268.2168.2167.0369.5399.3599.55
)], attaches=null, journal=Journal(id=1324759606947315908, delFlag=0, nameCn=计算机研究与发展, nameEn=Journal of Computer Research and Development, nameHistory1=null, nameHistory2=null, issn=1000-1239, eissn=null, cn=11-1777/TP, coden=null, periodic=0, language=CN, oaType=null, ccby=null, superviseOffice=null, ownerOffice=null, pubOffice=null, editorOffice=null, officeType=null, aims=null, clcCode=null, officeProv=null, officeCity=null, officeAddr=null, officeZip=null, officeEmail=null, officePhone=null, editDirector=null, officeDirector=null, officeDirectorPhone=null, officeStaffNum=null, officeEmpNum=null, coverPicUrl=k2l8zgqwstuxxevFX2qxVA==, journalPrice=null, startedYear=null, abbrevIsoEn=Journal of Computer Research and Development, journalRemark=null, publicationField=null, createdTime=1793776903141, updatedTime=1793850698612, createdBy=18614031015, updatedBy=13701087609, firstLetterCn=J, firstLetterEn=J, subjectCode=Electronic Information, subjectName=null, subjectCodeEn=Electronic Information, subjectNameEn=null, picCn=k2l8zgqwstuxxevFX2qxVA==, picEn=r6k0Ry+RD+c/VXj5QarmDw==, jcr=null, cjcr=null, exts=[JournalExt(id=1325069127710286688, language=CN, name=计算机研究与发展, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1793850698640, updatedTime=1793850698640, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://crad.ict.ac.cn/journalx/authorLogOn.action, submissionEditorUrl=https://crad.ict.ac.cn/journalx/editorLogOn.action, submissionReviewUrl=https://crad.ict.ac.cn/journalx/expertLogOn.action, submissionCeEditorUrl=https://crad.ict.ac.cn/journalx/editorInChiefLogOn.action, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1325069127777395553, language=EN, name=Journal of Computer Research and Development, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1793850698656, updatedTime=1793850698656, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://crad.ict.ac.cn/journalx/authorLogOn.action, submissionEditorUrl=https://crad.ict.ac.cn/journalx/editorLogOn.action, submissionReviewUrl=https://crad.ict.ac.cn/journalx/expertLogOn.action, submissionCeEditorUrl=https://crad.ict.ac.cn/journalx/editorInChiefLogOn.action, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1324759914146529478, websiteList=[Website(id=1325070300139892771, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1324759914146529478, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN, language=CN, createTime=1793850978169, createBy=18614031015, updateTime=1793851040259, updateBy=18614031015, name=计算机研究与发展-中文, tplId=1146099689490845704, title=计算机研究与发展, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1325070893751349360, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=articleTextType, value=kx, createTime=1793851119697, updateTime=1793851119697, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893730377837, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=banner, value=null, createTime=1793851119692, updateTime=1793851119692, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893772320883, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=grayFlag, value=0, createTime=1793851119702, updateTime=1793851119702, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893721989228, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=logo, value=https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN/file/pic?fileId=U27tDsy717YF7XSPA+Vwdw==, createTime=1793851119690, updateTime=1793851119690, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893784903797, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=minRunFlag, value=0, createTime=1793851119705, updateTime=1793851119705, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893742960751, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN/file/pic, createTime=1793851119695, updateTime=1793851119695, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893780709492, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=silenceFlag, value=0, createTime=1793851119704, updateTime=1793851119704, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893738766446, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1793851119694, updateTime=1793851119694, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893755543665, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=themeColor, value=null, createTime=1793851119698, updateTime=1793851119698, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893763932274, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=themeStyle, value=null, createTime=1793851119700, updateTime=1793851119700, creator=18614031015, updator=18614031015)]), Website(id=1325070300580294698, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1324759914146529478, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/jsjyjyfz/EN, language=EN, createTime=1793850978274, createBy=18614031015, updateTime=1793851036638, updateBy=18614031015, name=计算机研究与发展-英文, tplId=1146101810881728533, title=Journal of Computer Research and Development, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1325070925288321148, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=articleTextType, value=kx, createTime=1793851127216, updateTime=1793851127216, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925263155321, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=banner, value=null, createTime=1793851127210, updateTime=1793851127210, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925309292671, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=grayFlag, value=0, createTime=1793851127221, updateTime=1793851127221, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925254766712, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=logo, value=https://castjournals.cast.org.cn/joweb/jsjyjyfz/EN/file/pic?fileId=U27tDsy717YF7XSPA+Vwdw==, createTime=1793851127208, updateTime=1793851127208, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925326069889, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=minRunFlag, value=0, createTime=1793851127225, updateTime=1793851127225, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925279932539, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/jsjyjyfz/EN/file/pic, createTime=1793851127214, updateTime=1793851127214, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925317681280, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=silenceFlag, value=0, createTime=1793851127223, updateTime=1793851127223, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925271543930, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1793851127212, updateTime=1793851127212, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925292515453, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=themeColor, value=null, createTime=1793851127217, updateTime=1793851127217, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925300904062, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=themeStyle, value=null, createTime=1793851127219, updateTime=1793851127219, creator=18614031015, updator=18614031015)])], journalTitle=计算机研究与发展, weixinUrl=null, journalUrl=https://crad.ict.ac.cn/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Journal of Computer Research and Development, journalPhotoCn=k2l8zgqwstuxxevFX2qxVA==, journalPhotoEn=r6k0Ry+RD+c/VXj5QarmDw==, journalFirstLetter=J, journalRecommend=null, journalNew=null, journalCollection=null, jcrJf=null, cjcrJf=null, jcrJfStr=null, cjcrJfStr=null, submissionFirstDecision=null, sciSubjectClassification=null, casSubjectClassification=null, citeScore=null, totalCitationFrequency=null, icpCode=null, psCode=null, advertisingLicenseCode=null, copyrightInformation=null, country=null, option=, provinceCode=null, provinceName=null, collectFlag=false, interPubPlatform=, interPubPlatformUrl=null), detailUrlCn=https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN/10.7544/issn1000-1239.202550399, detailUrlEn=https://castjournals.cast.org.cn/joweb/jsjyjyfz/EN/10.7544/issn1000-1239.202550399, pdfUrlCn=https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN/PDF/10.7544/issn1000-1239.202550399, pdfUrlEn=https://castjournals.cast.org.cn/joweb/jsjyjyfz/EN/PDF/10.7544/issn1000-1239.202550399, aliStartDate=0, aliEndDate=0, collectionFlag=false, citedCount=null, citedUrl=null, previewStatus=0, delFlag=0, hasFullText=1, orderTime=1759248000000, fullTextJson=null, articleText=null, reference=null)
收藏切换
低时延智能体网络中基于三阶段关联验证的伪造链路攻击检测
张峰 , 俸皓 , 周秀民 , 章茂淳 , 陈镇铭 , 刘玉明 *
计算机研究与发展 | 新型分布式网络与系统中的数据安全与隐私保护专题 2025,62(10): 2495-2511
收起
收藏切换
计算机研究与发展 |新型分布式网络与系统中的数据安全与隐私保护专题 2025 , 62 (10) : 2495 -2511
低时延智能体网络中基于三阶段关联验证的伪造链路攻击检测
全屏
[Author(id=1327018984005808281, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=zhangfeng@guet.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1327018984068722843, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018984005808281, language=EN, stringName=Feng Zhang, firstName=Feng, middleName=null, lastName=Zhang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004, bio={"content":"

born in 1977. PhD. His main research interests include network security and data science

"}, bioImg=null, bioContent=

born in 1977. PhD. His main research interests include network security and data science

, aboutCorrespAuthor=null), CN=AuthorExt(id=1327018984148414620, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018984005808281, language=CN, stringName=张峰, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=桂林电子科技大学计算机与信息安全学院 广西桂林 541004, bio={"img":"QPZ8J4/RKKLmybxkwngHAQ==","content":"

1977年生. 博士. 主要研究方向为网络安全、数据科学

"}, bioImg=QPZ8J4/RKKLmybxkwngHAQ==, bioContent=

1977年生. 博士. 主要研究方向为网络安全、数据科学

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1327018983921922197, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, xref=null, ext=[AuthorCompanyExt(id=1327018983934505110, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004), AuthorCompanyExt(id=1327018983942893719, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=桂林电子科技大学计算机与信息安全学院 广西桂林 541004)])]), Author(id=1327018984207134878, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1327018984278438048, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018984207134878, language=EN, stringName=Hao Feng, firstName=Hao, middleName=null, lastName=Feng, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004, bio={"content":"

born in 1978. PhD, associate professor. His main research interests include computing intelligence and distributed system

"}, bioImg=null, bioContent=

born in 1978. PhD, associate professor. His main research interests include computing intelligence and distributed system

, aboutCorrespAuthor=null), CN=AuthorExt(id=1327018984442015905, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018984207134878, language=CN, stringName=俸皓, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=桂林电子科技大学计算机与信息安全学院 广西桂林 541004, bio={"img":"nolLLZSPlwth3t/lt3ihpQ==","content":"

1978年生. 博士,副教授. 主要研究方向为计算智能、分布式系统

"}, bioImg=nolLLZSPlwth3t/lt3ihpQ==, bioContent=

1978年生. 博士,副教授. 主要研究方向为计算智能、分布式系统

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1327018983921922197, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, xref=null, ext=[AuthorCompanyExt(id=1327018983934505110, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004), AuthorCompanyExt(id=1327018983942893719, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=桂林电子科技大学计算机与信息安全学院 广西桂林 541004)])]), Author(id=1327018984517513379, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1327018984597205157, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018984517513379, language=EN, stringName=Xiumin Zhou, firstName=Xiumin, middleName=null, lastName=Zhou, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004, bio={"content":"

born in 2000. Master candidate. His main research interests include agentic network and AIoT

"}, bioImg=null, bioContent=

born in 2000. Master candidate. His main research interests include agentic network and AIoT

, aboutCorrespAuthor=null), CN=AuthorExt(id=1327018984664314022, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018984517513379, language=CN, stringName=周秀民, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=桂林电子科技大学计算机与信息安全学院 广西桂林 541004, bio={"img":"VkVZyN6tP5P3iDaTB7Th5Q==","content":"

2000年生. 硕士研究生. 主要研究方向为智能体网络、人工智能物联网

"}, bioImg=VkVZyN6tP5P3iDaTB7Th5Q==, bioContent=

2000年生. 硕士研究生. 主要研究方向为智能体网络、人工智能物联网

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1327018983921922197, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, xref=null, ext=[AuthorCompanyExt(id=1327018983934505110, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004), AuthorCompanyExt(id=1327018983942893719, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=桂林电子科技大学计算机与信息安全学院 广西桂林 541004)])]), Author(id=1327018984752394408, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, orderNo=3, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1327018984857252010, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018984752394408, language=EN, stringName=Maochun Zhang, firstName=Maochun, middleName=null, lastName=Zhang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004, bio={"content":"

born in 2001. Master candidate. His main research interests include artificial intelligence and agentic network

"}, bioImg=null, bioContent=

born in 2001. Master candidate. His main research interests include artificial intelligence and agentic network

, aboutCorrespAuthor=null), CN=AuthorExt(id=1327018984928555179, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018984752394408, language=CN, stringName=章茂淳, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=桂林电子科技大学计算机与信息安全学院 广西桂林 541004, bio={"img":"igXoN2sl3+JHtOS8MRvcyQ==","content":"

2001年生. 硕士研究生. 主要研究方向为人工智能、智能体网络

"}, bioImg=igXoN2sl3+JHtOS8MRvcyQ==, bioContent=

2001年生. 硕士研究生. 主要研究方向为人工智能、智能体网络

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1327018983921922197, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, xref=null, ext=[AuthorCompanyExt(id=1327018983934505110, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004), AuthorCompanyExt(id=1327018983942893719, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=桂林电子科技大学计算机与信息安全学院 广西桂林 541004)])]), Author(id=1327018985062772909, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, orderNo=4, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1327018985352179887, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018985062772909, language=EN, stringName=Zhenming Chen, firstName=Zhenming, middleName=null, lastName=Chen, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004, bio={"content":"

born in 2000. Master candidate. His main research interests include artificial intelligence and agentic network

"}, bioImg=null, bioContent=

born in 2000. Master candidate. His main research interests include artificial intelligence and agentic network

, aboutCorrespAuthor=null), CN=AuthorExt(id=1327018985419288752, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018985062772909, language=CN, stringName=陈镇铭, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=桂林电子科技大学计算机与信息安全学院 广西桂林 541004, bio={"img":"zot74ZXaSynezrhN0vft2Q==","content":"

2000年生. 硕士研究生. 主要研究方向为人工智能、智能体网络

"}, bioImg=zot74ZXaSynezrhN0vft2Q==, bioContent=

2000年生. 硕士研究生. 主要研究方向为人工智能、智能体网络

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1327018983921922197, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, xref=null, ext=[AuthorCompanyExt(id=1327018983934505110, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004), AuthorCompanyExt(id=1327018983942893719, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=桂林电子科技大学计算机与信息安全学院 广西桂林 541004)])]), Author(id=1327018985494786226, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, orderNo=5, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=ymliu@guet.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1327018985645781172, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018985494786226, language=EN, stringName=Yuming Liu, firstName=Yuming, middleName=null, lastName=Liu, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004, bio={"content":"

born in 1991. PhD. His main research interests include software-defined networking, programmable networks, and network security

"}, bioImg=null, bioContent=

born in 1991. PhD. His main research interests include software-defined networking, programmable networks, and network security

, aboutCorrespAuthor=null), CN=AuthorExt(id=1327018985721278645, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, authorId=1327018985494786226, language=CN, stringName=刘玉明, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=桂林电子科技大学计算机与信息安全学院 广西桂林 541004, bio={"img":"lKp8Wv0iiRUbZkApFKp6+A==","content":"

1991年生. 博士. 主要研究方向为软件定义网络、可编程网络、网络安全

"}, bioImg=lKp8Wv0iiRUbZkApFKp6+A==, bioContent=

1991年生. 博士. 主要研究方向为软件定义网络、可编程网络、网络安全

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1327018983921922197, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, xref=null, ext=[AuthorCompanyExt(id=1327018983934505110, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004), AuthorCompanyExt(id=1327018983942893719, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1327018220717978265, companyId=1327018983921922197, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=桂林电子科技大学计算机与信息安全学院 广西桂林 541004)])])]
张峰 , 俸皓, 周秀民, 章茂淳, 陈镇铭, 刘玉明*
作者信息
  • 桂林电子科技大学计算机与信息安全学院 广西桂林 541004
通讯作者:
刘玉明()
作者简介:

1977年生. 博士. 主要研究方向为网络安全、数据科学

1978年生. 博士,副教授. 主要研究方向为计算智能、分布式系统

2000年生. 硕士研究生. 主要研究方向为智能体网络、人工智能物联网

2001年生. 硕士研究生. 主要研究方向为人工智能、智能体网络

2000年生. 硕士研究生. 主要研究方向为人工智能、智能体网络

1991年生. 博士. 主要研究方向为软件定义网络、可编程网络、网络安全

Tri-Phase Correlation Verification Based Fabricated Link Attacks Detection in Low-Latency Agentic Networks
Feng Zhang , Hao Feng, Xiumin Zhou, Maochun Zhang, Zhenming Chen, Yuming Liu*
Affiliations
  • School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004
  • born in 1977. PhD. His main research interests include network security and data science

    born in 1978. PhD, associate professor. His main research interests include computing intelligence and distributed system

    born in 2000. Master candidate. His main research interests include agentic network and AIoT

    born in 2001. Master candidate. His main research interests include artificial intelligence and agentic network

    born in 2000. Master candidate. His main research interests include artificial intelligence and agentic network

    born in 1991. PhD. His main research interests include software-defined networking, programmable networks, and network security

出版时间: 2025-10-01 doi: 10.7544/issn1000-1239.202550399
文章导航
收藏切换

面向低时延的智能体网络环境,研究了一种基于关联链路验证的伪造链路攻击检测方法. 在该网络环境下,正常链路与伪造链路在测量时延等关键指标上的统计特征差异较小,并且背景流量和协调器过载攻击的影响会放大测量误差,但缩小该差异会导致现有方法的检测性能骤降. 对此,提出一种基于关联链路验证的伪造链路攻击检测方法,该方法包含3个阶段:首先,采用一种有效时延转换方法缓解测量误差的影响,并将处理后的测量时延等转化为链路性能指标;然后,根据提出的多径传输模拟方法,将关联链路之间的性能差异转化为易于观测的统计特征;最后,基于极值理论和概率分布拟合方法确定统计特征的阈值,用于伪造链路的检测. 仿真实验结果表明,在不同的网络规模和攻击场景下,所提出的方法能够有效地检测低时延智能体网络中的伪造链路攻击,并且在检测性能方面明显优于现有的相关方法,F1分数提升达20%~30%. 特别地,当链路传输时延低至0.1 ms时,仍然能取得95.23%的检测率和0.28%的误判率,显示出良好的鲁棒性.

智能体网络  /  低时延  /  伪造链路攻击  /  关联链路验证  /  极值理论

The critical challenge of detecting fabricated link attacks in low-latency agentic network environments, where conventional detection methods face significant limitations is investigated in this paper. The fundamental difficulty stems from three key factors: 1) the measured performance differences between normal and fabricated links often exhibit marginal variations that fall below meaningful detection thresholds; 2) dynamic background traffic patterns introduce substantial interference that obscures genuine attack signatures; and 3) coordinated overload attacks deliberately distort measurement accuracy. These factors collectively degrade the detection performance of existing methods by significantly reducing the observable distinction between legitimate and malicious links. To address these challenges, we propose a novel three-phase detection framework based on correlated link verification. The first phase implements an effective delay transformation method that reduces measurement errors while converting raw measurements into meaningful link performance metrics. The second phase employs a multipath transmission simulation technique that amplifies subtle performance disparities between correlated links into statistically observable features. The final phase establishes robust detection thresholds through the combined application of extreme value theory and probability distribution fitting methods. Comprehensive simulation results validate that our proposed method consistently outperforms existing approaches across diverse network scales and attack scenarios. The framework demonstrates particular effectiveness in low-latency agentic environments, maintaining superior detection performance while meeting stringent timing requirements. Specifically, the framework achieves a 20% to 30% improvement in F1 score compared with state-of-the-art methods. Notably, even under ultra-low latency conditions (e.g., 0.1ms link delay), the proposed method maintains a high detection rate of 95.23% with a false positive rate as low as 0.28%. These advancements provide a practical solution for securing next-generation autonomous networks against sophisticated link fabrication attacks.

agentic network  /  low-latency  /  fabricated link attacks  /  correlated link verification  /  extreme value theory
张峰, 俸皓, 周秀民, 章茂淳, 陈镇铭, 刘玉明. 低时延智能体网络中基于三阶段关联验证的伪造链路攻击检测. 计算机研究与发展, 2025 , 62 (10) : 2495 -2511 . DOI: 10.7544/issn1000-1239.202550399
Feng Zhang, Hao Feng, Xiumin Zhou, Maochun Zhang, Zhenming Chen, Yuming Liu. Tri-Phase Correlation Verification Based Fabricated Link Attacks Detection in Low-Latency Agentic Networks[J]. Journal of Computer Research and Development, 2025 , 62 (10) : 2495 -2511 . DOI: 10.7544/issn1000-1239.202550399
随着多智能体系统(multi-agent systems,MAS)与人工智能服务的持续融合,智能体网络(agentic network)逐渐发展成为一种面向智能体自治协同的新型网络范式,并已在智能工厂、智慧城市等多种AIoT场景中得到广泛应用[1-2]. 该架构旨在支持异构智能体在跨域环境中实现高效的协同感知、推理决策与服务执行,具有高度自主性和去中心化的特征[3]. 在实际部署中,智能体网络通常通过统一的注册机制与发现协议,将大量分布式智能体接入网络,并为其分配唯一的数字身份. 这一过程多基于应用层协议,如agent network protocol(ANP),完成智能体的注册、认证、能力通告及调度协作等功能[4].
尽管上述逻辑层机制在一定程度上提升了智能体间交互的可信性与协同性,智能体网络的安全性仍面临多维威胁. 根据攻击行为的作用层级,可将其大致划分为2类:一类是链路层攻击,主要针对网络底层的物理与逻辑连接关系,典型代表如伪造链路、链路劫持等;另一类则是应用层攻击,例如恶意智能体注册、能力伪报、任务拒绝响应等行为,直接破坏调度协作的信任基础.
目前已有的安全研究多集中于应用层,重点关注智能体行为验证、调度策略鲁棒性和恶意智能体识别等方向[5],而对链路层安全问题的研究相对不足. 当前应用层的安全主要通过身份管理、加密、认证和授权等机制进行保证,交互过程通常以秒为单位,考虑到智能体网络中智能体的动态特性(可能频繁地加入、离开网络),在以智能工厂为代表的需要以毫秒为单位的典型低延时应用场景中,这些方法将不再适用. 本文正是聚焦于这一被忽视的层面,研究其潜在风险与防御方法. 由于智能体网络在底层通信中仍依赖传统网络设施的链路结构,尤其在智能体初次接入或协作路径更新过程中,控制面需通过链路发现机制掌握当前的物理拓扑. 这一过程极易遭受利用拓扑发现协议设计缺陷发起的伪造链路攻击. 以软件定义网络(software-defined networking,SDN)中的链路层发现协议(如 LLDP)为例,已有研究表明攻击者可以通过中继或伪造探测报文,诱使控制器误判2个物理上并未连接的节点之间存在链路[6]. 类似方式在智能体网络中同样适用,可能导致协调器构建出错误的逻辑拓扑视图,进而引发一系列严重后果[7-8]. 一方面,路径选择与资源调度将在不可信拓扑信息基础上执行,导致通信延迟增加和任务分配失衡;另一方面,推理数据路径被劫持或重构,可能破坏感知输入与语义一致性,降低系统推理精度与稳定性. 更为严重的是,攻击者可借助伪造链路将受控智能体引入协同推理过程,操纵中间结果、误导模型更新,甚至植入后门知识,从而将智能体网络由自治协作系统演变为可被远程操控的攻击工具,对关键基础设施构成长期隐蔽威胁.
这一问题在面向智能制造、车联网等对低时延与可靠性要求极高的 AIoT 场景中更为突出. 在这类场景中,智能体之间的协同感知、任务分派与执行反馈高度依赖于底层链路的时延与稳定性. 因此,链路质量的准确感知成为控制面进行调度与路径规划的关键基础. 一旦攻击者利用控制器在处理探测报文时产生的“抖动”误差,构造出具有真实延迟特征的伪造路径,便可有效绕过基于延迟分布建模或最短路径一致性分析的传统检测机制[9-10]. 此外,由于伪造链路攻击不会显著增加网络流量或引起通信负载异常,现有依赖流量特征或行为模式的检测方法难以奏效[11-12].
为应对上述挑战,本文提出了一种面向智能体网络的链路层攻击检测方法——CLV(correlated link verification). 该方法基于“关联链路”概念,评估由同一智能体发起、处于相似网络条件下的多个邻接链路在通信时延特性上的一致性,从而识别出存在异常偏差的可疑伪造路径. CLV方法具备部署轻量、信息依赖低等优势,即无需引入额外控制消息,不依赖高维行为建模,仅通过简单的通信延迟测量即可实现高效且可扩展的拓扑安全检测.
针对伪造链路的检测,目前主流的检测方法为静态或动态的链路时延检测方法,它们的主要区别在于是否更新检测阈值. 静态方法需在系统运行前进行部署,以采集全网或部分链路的时延作为基线数据集,并以此构建基于统计或机器学习的判定模型. 在系统运行后,根据采集的链路数据判定新增链路是否是伪造的. 动态方法在系统运行中即可部署,通过实时采集的链路时延更新检测阈值,以对新增的链路的合法性进行判定. 另外,还有一些检测方法需要对攻击场景进行特定的假设,因此具有一定的局限性.
在静态的检测方法中,Smyth等人[13]较早地关注了SDN环境下的伪造链路攻击,通过分析不同中继节点数量以及中继模块处理能力情况下伪造链路时延与正常链路时延分布的差异,提出了基于p值的检测方法. 为增强检测方法的可扩展性,Al-Salti等人[14]设计了LINK-GUARD,通过采集不同网络规模下的链路时延来构建数据集,并基于箱线图方法检测异常链路. 然而,由于未考虑网络波动与控制器扰动的影响,这些方法缺乏鲁棒性. 为此,Soltani等人[15]设计了MLLG,它不仅考虑了交换机之间的链路时延,也考虑了交换机与控制器之间的链路时延. 利用这些时延构建数据集,并采用机器学习的方法训练判别模型. 进一步地,Soltani等人[16]设计了RLV,通过采集不同网络规模、不同伪造链路数量和是否包含引入背景流量等情景下的链路时延,以训练更具扩展性和鲁棒性的判别模型. 然而,在低时延的网络环境下,网络波动与控制器扰动对测量结果的干扰尤为明显,这些方法的检测效果仍然不够理想,难以有效地发现伪造链路攻击.
在动态的检测方法中,Skowyra等人[17]设计了TopoGuard+,使用全局队列存储采集到的链路时延,并根据定义的阈值计算方法动态地更新检测阈值. 若链路的测量时延超过阈值,则认为该链路是伪造的,随后启用防御模块短暂地停用该链路. 但如文献[18]所述,TopoGuard+中的检测方法存在缺陷. 攻击者可通过扰动控制器来增大测量误差和检测阈值,进而绕过检测. Chou等人[19]和Huang等人[20]分别设计了CTAD和TrustTopo,通过修正阈值的计算方法,来降低阈值的变化幅度,在一定程度上提高了检测率. 然而,上述动态检测的方法均难以设计合适的阈值,以准确地区分正常链路和伪造链路,仍然缺乏鲁棒性.
此外,还有一些基于特定假设的检测方法. Shrivastava等人[21]设计了SRD,在假设中继模块无法处理巨帧的前提下,利用巨型探测报文探测发现伪造链路. Wang等人[22]假设伪造链路攻击会引起大量会话中断,并提出基于强化学习的检测方法. Kong等人[23]提出一种具有攻击维持能力的伪造链路攻击模型,能够突破现有方法的检测,并对此提出一种基于转发路径验证的检测方法. 这些方法对攻击者的能力进行了明确限制,缺乏普适性. 此外,由于伪造链路攻击本身不破坏传输报文的完整性,一些基于安全协议[24-28]的拓扑欺骗防御方法,以及针对主机位置欺骗攻击的防御方法[29-31]将无法适用.
智能体网络中探测数据包的转发过程如图1(a)所示. 首先,智能体协调器构造探测数据包,将其发送至各Agent worker(步骤①). 接着,Agent worker收到探测数据包后,通过广播的形式将其转发至与之连接的其他Agent worker(步骤②). 当某个Agent worker收到来自其邻居的探测数据包后,将其转发至Agent协调器(步骤③). 最后,Agent协调器根据闭环的探测数据包转发路径,确定Agent worker之间的连接,进而构建全局的拓扑视图.
然而,上述链路发现过程可能受到伪造链路攻击的威胁. 由于探测数据包广播传递,恶意Agent worker能够在正常Agent worker之间中继该消息,使Agent协调器误以为正常Agent worker之间存在一条链路,该链路实际由恶意Agent worker伪造. 具体流程如图1(b)中步骤③~⑥所示. 基于该伪造链路,恶意Agent worker可能丢弃、监听、篡改正常Agent worker之间的业务通信数据.
尽管有研究工作通过探测数据包测量链路时延,并根据异常的链路时延值发现伪造链路,但它们难以适用于低时延的网络环境. 在低时延场景下,Agent协调器测量的链路时延容易受到背景流量等因素的影响. 不仅如此,恶意Agent worker可能向Agent协调器发起过载攻击,发送大量查询请求增加其负载,影响链路时延的测量结果,如图1(b)中的步骤①②所示. 在此背景下,链路时延测量误差被放大,导致现有方法检测性能急剧下降,难以准确检测伪造链路的存在.
本节首先建立测量的链路时延的模型,分析伪造链路攻击的特点以及检测该攻击存在的挑战. 接着,给出现有方法的形式化描述,分析它们在低时延网络中难以准确识别伪造链路的原因. 本节主要使用的符号及其含义如表1所示.
根据图1可知,正常链路的测量时延$ {T_{\rm nor}} $和伪造链路的测量时延$ {T_{\rm mal}} $分别如式(1)和式(2)所示.
$ {T_{\rm nor}} = {T_{\rm trx}} + T_{\rm fwd}^{\rm s} + {T_{\rm err}} \text{,} $
$ {T_{\rm mal}} = \left( {m + 1} \right){T_{\rm trx}} + \sum\limits_{i = 1}^m {T_{\rm fwd}^{{\rm h},i} + {T_{\rm err}}} . $
其中,${T_{\rm trx}}$表示链路传输时延,$T_{\rm fwd}^{\rm s}$和$T_{\rm fwd}^{\rm h}$分别表示Agent worker端口和中继模块的转发时延,${T_{\rm err}}$表示测量误差,$m$表示中继模块的数量. 由于传播时延非常小,故此处将其忽略. 为便于分析,假设${T_{\rm trx}}$为常量$c$,${T_{\rm err}}$的分布未知. 假设报文按速率为${\lambda _{\rm lnk}}$的泊松过程到达,Agent worker端口和中继模块分别按速率为$ \lambda _{\rm fwd}^{\rm s} $和$ \lambda _{\rm fwd}^{\rm h} $的泊松过程转发. 通常情况下,中继模块相较于Agent worker端口的转发速率更低,在链路不过载的情况下有${\lambda _{\rm lnk}} < \lambda _{\rm fwd}^{\rm h} < \lambda _{\rm fwd}^{\rm s}$,此时Agent worker端口和中继模块的平均转发时延分别为${1 / { ( {\lambda _{\rm fwd}^{\rm s} - {\lambda _{\rm lnk}}} )}}$和$1 / ( \lambda _{\rm fwd}^{\rm h} - {\lambda _{\rm lnk}} )$. 因此,正常链路和伪造链路的测量时延的期望分别如式(3)和式(4)所示.
$ E\left( {{T_{\rm nor}}} \right) = c + \frac{1}{{\lambda _{\rm fwd}^{\rm s} - {\lambda _{\rm lnk}}}} + E\left( {{T_{\rm err}}} \right) \text{,} $
$ E\left( {{T_{\rm mal}}} \right) = \left( {m + 1} \right)c + \sum\limits_{i = 1}^m {\frac{1}{{\lambda _{\rm fwd}^{\rm h} - {\lambda _{\rm lnk}}}}} + \frac{1}{{\lambda _{\rm fwd}^{\rm s} - {\lambda _{\rm lnk}}}} + E\left( {{T_{\rm err}}} \right). $
通过对比式(3)(4)可知,$\varDelta = E\left( {{T_{\rm mal}}} \right) - E\left( {{T_{\rm nor}}} \right) > 0$,这表明正常链路与伪造链路之间存在固有的性能差异,这种差异性正是相关方法检测伪造链路的基础. 当满足如下3种情况时,差异$\varDelta $是较为显著的:1)传输时延较大,即$ c $取值较大时的情况. 2)中继模块的性能较低,即$ \lambda _{\rm fwd}^{\rm h} $取值较小时的情况. 此时,中继模块的转发时延较高. 3)伪造链路过载,即$\lambda _{\rm fwd}^{\rm h} < {\lambda _{\rm lnk}}$时的情况. 此时式(4)不再成立,并且转发时延会显著增加. 然而,在低时延网络中,攻击场景可能不再满足上述3种情况. 相应地,现有的检测方法难以准确地识别伪造链路. 根据阈值是否动态更新,现有方法可分为基于静态阈值和动态阈值的检测方法,其形式化描述和脆弱性分析如下.
静态检测方法的思想在于,通过采集的数据集训练检测模型,以识别伪造链路,其形式化描述如式(5)所示.
$ {Z_{\rm sta}}\left( {{l_i}} \right) = I\left( {\sum\limits_{t=1} {{g_{\rm tnt}}\left( {{D_{\rm sta}},\boldsymbol v_{{l_i}}^t} \right)} \geqslant {\delta _{\rm n}}} \right) , $
其中,$ {D_{\rm sta}} = \{ { \{ {\boldsymbol v_{{l_1}}^1,\boldsymbol v_{{l_2}}^1, … } \}, \{ {\boldsymbol v_{{l_1}}^2,\boldsymbol v_{{l_2}}^2, … } \}, … } \} $表示在检测模型部署前采集的数据集,用于模型的训练. $\boldsymbol v_{{l_i}}^t$表示链路${l_i}$在第$t$个采样周期的特征向量,用于模型的测试. ${g_{\rm tnt}}$表示训练和测试函数,它根据$ {D_{\rm sta}} $和$\boldsymbol v_{{l_i}}^t$得到一次判别结果. ${Z_{\rm sta}}$表示决策函数,根据累计的判别结果与阈值$ {\delta _{\rm n}} $的关系得到最终的检测结果. 不同的静态检测方法所使用的数据集、检测模型以及决策阈值各不相同. 在文献[13]中,$ \boldsymbol v_{{l_i}}^t = ({d_{{l_1}}^t}, {d_{{l_2}}^t},… ,{d_{{l_i}}^t} ),{g_{\rm tnt}} = p - value, {\delta _{\rm n}} > 1 $,其中$ d_{{l_i}}^t $表示在第$t$个采样周期链路${l_i}$的测量时延. 在文献[15]中,$ \boldsymbol v_{{l_i}}^t = ( {d_{{l_i}}^t,d_{\rm cs}^t,d_{\rm sc}^t} ),{g_{\rm tnt}} = ML,{\delta _{\rm n}} = 1 $,其中$ d_{\rm cs}^t $和$ d_{\rm sc}^t $分别表示控制器到交换机以及交换机到控制器的测量时延,$ ML $表示机器学习的模型. 在文献[14]中,$ \boldsymbol v_{{l_i}}^t = ({d_{{l_1}}^t}, {d_{{l_2}}^t},… ,{d_{{l_i}}^t}) ,{g_{\rm tnt}} = Q3 + 3 \times IQR,{\delta _{\rm n}} = 10 $,其中$Q3$和$IQR$分别表示测量时延集合的第3个四分位数和四分位数间距.
动态检测方法的思想在于,通过采集的数据集和动态的时延阈值来识别伪造链路,其统一的形式化描述如式(6)所示.
$ {Z_{\rm dyn}}\left( {{l_i}} \right) = I\left( {\sum\limits_{t=1} {I\left( {d_{{l_i}}^t > \delta _{\rm d}^t} \right)} \geqslant {\delta _{\rm n}}} \right) , $
其中,$ d_{{l_i}}^t $表示在第$t$个采样周期链路${l_i}$的测量时延,$\delta _{\rm d}^t$表示动态的时延阈值. ${Z_{\rm dyn}}$表示决策函数,它根据测量时延$ d_{{l_i}}^t $和时延阈值$\delta _{\rm d}^t$之间的大小关系得到一次判别结果,并根据累计的判别结果与决策阈值${\delta _{\rm n}}$得到最终的检测结果. 时延阈值$\delta _{\rm d}^t$通过$\delta _{\rm d}^t = {g_{\rm thr}} ( {{D_{\rm dyn}}} )$进行计算. 其中,${g_{\rm thr}}$为时延阈值的计算函数,$ {D_{\rm dyn}} = \{ \{ d_{{l_1}}^1, d_{{l_2}}^1, … \}, … \{ {d_{{l_1}}^t,d_{{l_2}}^t, … } \} \} $表示截至第$t$个采样周期采集到的数据集. 不同的动态检测方法使用的阈值计算函数和决策阈值各不相同. 在文献[17]中,${g_{\rm thr}} = Q3 + 3 \times IQR,{\delta _{\rm n}} = 1$. 在文献[19]中,${g_{\rm thr}} = Q3 + 1.5 \times IQR, {\delta _{\rm n}} = 1$. 在文献[20]中,${g_{\rm thr}} = {{\bar d_{{l_i}}}} + \alpha \max ( d_{{l_i}}^t - {{\bar d_{{l_i}}}} , 0 ) + c,{\delta _{\rm n}} \geqslant 1$,其中$ {{\bar d_{{l_i}}}} $表示平均时延,$\alpha $表示测量结果超过均值$ {{\bar d_{{l_i}}}} $的比例,$c$为修正阈值的常数项.
最后,分析这些检测方法的脆弱性. 由式(5)可知,静态检测方法的检测性能主要取决于数据集$ {D_{\rm sta}} $的质量. 然而,在低时延网络中测量误差是不容忽视的,它可能污染数据集和削弱模型的检测效果. 另外,静态检测方法适应性较差,针对一个环境的数据集和模型难以直接应用于其他环境,相比之下动态检测方法适应性更强. 由式(6)可知,动态检测方法的检测性能取决于动态阈值$ \delta _{\rm d}^t $的选取. 然而,现有方法直接根据测量结果的数值大小或统计特征来确定阈值缺乏鲁棒性. 攻击者可通过发起过载攻击增大测量误差,进而增大阈值$ \delta _{\rm d}^t $,使得$ d_{{l_i}}^t < \delta _{\rm d}^t $来规避检测[15-16,18]. 不仅如此,攻击者可通过优化中继模块的性能来降低转发时延,缩小伪造链路与正常链路之间的性能差异. 此时,确定合适的$ \delta _{\rm d}^t $是有挑战的,$ \delta _{\rm d}^t $过高会导致较低的检测率,而过低则会导致较高的误报率.
如图2所示,关联链路指从同一Agent worker出行的2条处理探测数据包的链路,考虑到广播的探测数据包几乎在相同时刻沿关联链路发出,正常情况下2条链路的测量结果差异是相对确定且较小的. 基于该思路,CLV检测方法的工作流程如图3所示. 需要说明的是,图2中展示的网络拓扑仅为便于阐述关联链路的概念而进行的简化表示,并不意味着各Agent worker必须仅连接2个邻居. 实际上,其与任意2个邻居节点之间的出行链路均可构成一组关联链路.
在攻击检测之前,通过提出的数据处理方法,将采集的原始数据经过处理后用来转化链路质量,目的是缓解测量误差,以准确地反映实际的链路性能,详见3.2节的内容. 在攻击检测阶段,CLV首先通过提出的多径传输模拟方法,将测量的关联链路之间的性能差异转化为易于观测的统计特征,详见3.3节的内容. 接着,根据历史的统计特征序列,基于极值理论和概率分布拟合的方法确定统计特征的阈值,详见3.4节的内容. 最后,获取关联链路当前时刻的统计特征,并结合确定的阈值识别其中的伪造链路.
为评估链路的实时性能,引入链路质量的概念. 根据式(4)可知,链路时延和可承受负载可作为链路质量的评估指标,且低时延与高负载均能指示更优的链路质量. 需要说明的是,在低时延网络中,高时延表明测量误差是显著的,故仅当测量的链路时延较低时,链路负载才能有效地指示链路质量的优劣. 因此,给出链路质量的定义如式(7)所示. 表示在第$t$个采样周期链路的时延评分如式(8)所示. 表示链路的负载评分如式(9)所示.
$ w_l^t = \frac{{w_{{\rm d},l}^t\left( {1 + \exp \left( {w_{{\rm d},l}^t - 1} \right) \times w_{\lambda ,l}^t} \right)}}{2} \text{,} $
$ w_{{\rm d},l}^t = 1 - \tanh \left( {\frac{{\widetilde {d_l^t}}}{{{d_{\rm lim}}}}} \right) \text{,} $
$ w_{\lambda ,l}^t = \frac{1}{{1 + \exp \left( {\alpha {\lambda _{\max }} - \widetilde {\lambda _l^t}} \right)}} . $
在式(8)中,$d_l^t$表示在第$t$个采样周期链路$l$的测量时延,${d_{\rm lim}}$表示有效时延的上界,$ \widetilde {d_l^t} $表示经过转化的有效时延. 其意义在于,${d_{\rm lim}}$用于指示测量时延的有效性,即仅当测量时延低于该值时被认为有效,否则该测量时延被认为不能有效反映实际链路性能. $ \widetilde {d_l^t} $根据算法1和输入$d_l^t$获得. 具体地,若当前测量时延有效,则通过移动平均的方法得到最近几个采样周期的平均时延,如行⑨~⑪所示. 反之,继续判断有效时延队列的长度是否达到统计要求,若达到则取有效时延队列的中值,否则取真实时延队列的中值,如行③~⑧所示.
算法1借鉴了缺失值的处理思想,将超过${d_{\rm lim}}$的测量时延值视作缺失值. 由于时延评分对于低时延值更加敏感,故采用$\tanh $函数归一化时延评分. 在式(9)中,$ \lambda _l^t $表示在第$t$个采样周期链路$l$的测量负载. $ \widetilde {\lambda _l^t} $表示经过转化的有效负载,转化的方法与算法1一致. $ {\lambda _{\max }} $表示链路带宽,当$ 0 < \alpha \leqslant 1 $时,$ \alpha {\lambda _{\max }} $表示链路是否过载的界限. 由于当链路负载超过$ \alpha {\lambda _{\max }} $,转发时延会陡然增加,负载评分需相应地快速降低,因此采用${\rm sigmoid} $函数归一化负载评分.
算法1. 有效时延转化算法.
输入:测量的链路时延$ d_l^t $,有效时延的限值$ {d_{\rm lim}} $;
输出:经过转化的有效链路时延$ \widetilde {d_l^t} $.
① 若系统处于初始状态,则初始化有效时延 队列$valid\_latency\_queue$和真实时延$real\_$ $latency\_queue $为空;
② 将$ d_l^t $添加至循环队列$real\_latency\_queue$;
③ if $ d_l^t > {d_{\rm lim}} $ then
④ if $valid\_latency\_queue$长度不够then
⑤   $ \widetilde {d_l^t} \leftarrow median\left( {real\_latency\_queue} \right) $;
⑥ else
⑦   $ \widetilde {d_l^t} \leftarrow median\left( {valid\_latency\_queue} \right) $;
⑧ end if
⑨ else
⑩ 将$ d_l^t $添加至循环队列$valid\_latency\_queue$;
⑪  $ \widetilde {d_l^t} \leftarrow moving\_average\left( {valid\_latency\_queue} \right) $;
⑫ end if
为便于用统计特征刻画关联链路之间的性能差异,考虑将关联链路上每个周期的采样结果转化为离散的序列,本节给出一种多径传输模拟(multipath transmission simulation,MTS)方法,如算法2所示. 其原理如下:引入一个虚拟的会话,根据关联链路上每个周期的采样结果,选择其中性能更优的一条链路传输一个分组. 经过多个采样周期后,虚拟会话的选择结果能够反映链路之间的性能差异. 同时,虚拟会话的选择结果可以通过0和1组成的轨迹序列表达,其中1和0分别表示是否选择该链路. 因此,通过分析轨迹序列的统计特征,即可评估关联链路之间的性能差异.
算法2. 多径传输模拟算法.
输入:关联链路$corr\_links$,链路质量$ w_l^t $,模拟步长$steps$,折扣率${\gamma _0}$,统计周期${T_{\rm sta}}$,检测周期${T_{\rm det}}$;
输出:统计特征$K$和$H$的序列$ {K_{\rm seq}} $和$ {H_{\rm seq}} $.
① 初始化列表$ T{r_{\rm seq}} $,$ {K_{\rm seq}} $,$ {H_{\rm seq}} $为空;
② for $t{\text{ from 0 to }}steps$ do
③ for $l \in corr\_links$ do
④  if $ w_l^t \times \gamma _l^t $更大then
⑤   将$ l.id $添加至轨迹序列$ T{r_{\rm seq}} $;/*从1开始    的0-1序列*/
⑥    $\gamma _l^{t + 1} \leftarrow \gamma _l^t \times {\gamma _0}$;
⑦  else
⑧    $\gamma _l^{t + 1} \leftarrow 1$;
⑨  end if
⑩ end for
⑪ end for
⑫ 分别按照统计周期${T_{\rm sta}}$和检测周期$T_{\rm det} $,将$ T{r_{\rm seq}} $ 划分为多个子序列的集合$Tr_{\rm{s}eq}^{\mathrm{k}} $和$ Tr_{\rm{s}eq}^{\mathrm{h}} $;
⑬ for $ seq\in Tr_{\rm{s}eq}^{\mathrm{k}} $ do
⑭ 将子序列$ seq $中连续1的数量K添加至$K_{\rm seq} $;
⑮ end for
⑯ for $ seq\in Tr_{\rm{s}eq}^{\mathrm{h}} $
⑰  将子序列seq中0和1的分布的熵值H添加 至$H_{\rm seq} $;
⑱ end for
在算法2中,轨迹序列的获取过程如行④⑤所示. 需要说明的是,正常链路之间仍存在的十分细微的性能差异,若虚拟会话直接根据链路质量选择链路,则可能始终选择其中一条性能略优的链路,导致选择结果难以反映链路之间的性能差异. 为屏蔽链路之间这种细微的性能差异,算法2引入折扣因子来抑制虚拟会话连续选择同一条链路的行为. 具体地,虚拟会话根据链路质量与相应折扣因子的乘积来选择链路,当连续选择同一条链路时,折扣因子不断递减. 折扣因子的定义如式(10)所示,其更新过程如行⑥~⑧所示. 其中,$\gamma _l^t$表示在第$t$个采样周期链路的折扣因子,${\gamma _0}$表示折扣率,$b_l^t$表示在第$t$个采样周期链路是否被虚拟会话所选择.
$ \gamma _l^t = \left\{ \begin{aligned}& 1,\;\; b_l^tb_l^{t - 1} = 0 \text{,} \\& {\gamma _0}\gamma _l^{t - 1},\;\; b_l^tb_l^{t - 1} = 1 \text{,} \\ \end{aligned} \right. $
$ {T_{\rm sta}} = {T_{\rm spl}} \times \sum\limits_{l \in L} {\mathop {\max }\limits_{\prod\limits_{k = 1}^K {b_l^k = 1} } K} . $
统计特征$K$和$H$的计算说明如行⑭和行⑰所示,其含义解释如下:记关联链路$L = \{ {{l_i},{l_j}} \}$的轨迹序列为${S_{{l_i}}} = \left[ {b_{{l_i}}^0,b_{{l_i}}^1, … } \right]$,根据先前的分析,正常关联链路之间的性能差异十分细微. 因此,${S_{{l_i}}}$中0和1的数量应非常接近. 另外,折扣因子的引入旨在消除链路之间细微的性能差异,从而限制虚拟会话连续地选择同一条链路,故${S_{{l_i}}}$中连续1的长度较短,如${S_{{l_i}}} = [ 1,0,1,0,1,1, 0 ]$. 然而,伪造链路的存在使得虚拟会话倾向于选择性能更优的正常链路. 相应地,正常链路的轨迹序列中1的数量较多且连续1的长度较长,如${S_{{l_i}}} = [ 1,1, 1,1,1,0,1 ]$. 因此,利用香农熵$H$评估轨迹序列中0和1数量的均衡程度,而令$K$表示轨迹序列中连续1的长度. 另外,统计特征$H$和$K$被添加至列表中,用于后续阈值确定和伪造链路检测.
在算法2行⑫中,轨迹序列按照统计周期和检测周期被分为多个子序列,以便于统计特征的计算. 为了获取一个统计周期里连续1的数量,定义统计周期${T_{\rm sta}}$如式(11)所示,其中${T_{\rm spl}}$表示采样周期. 可见,统计周期${T_{\rm sta}}$的大小是动态的,它的取值取决于统计特征$K$的大小. 另外,为了获得稳定的结果,计算多个统计周期内的熵值,即${T_{\rm det}} = Interval \times {T_{\rm sta}}$,此处${T_{\rm det}}$表示计算熵值的周期.
为能够确定上述统计特征的阈值,提出一种动态阈值计算(dynamic threshold calculation,DTC)方法. 根据3.3节的分析可知,动态阈值应满足2个条件:1)当利用统计特征识别伪造链路时,阈值应使得式(12)成立. 2)当利用统计特征识别伪造链路时,阈值应使得式(13)成立. 需要说明的是,阈值的确定过程即为CLV方法的训练过程,而式(12)(13)中示性函数的结果即为CLV方法的检测结果,它们均能够指示伪造链路的存在.
$ {I_{\rm h}} = \left\{ \begin{aligned}1,\;\; H \leqslant {\delta _{\rm h}} \text{,} \\0,\;\; H > {\delta _{\rm h}}. \\ \end{aligned} \right. $
$ {I_{\rm k}} = \left\{ \begin{aligned}1,\;\; K > {\delta _{\rm k}} \text{,} \\0,\;\; K \leqslant {\delta _{\rm k}}. \\ \end{aligned} \right. $
首先,确定阈值${\delta _{\rm h}}$. 记熵值随机变量构成的序列为$[ {H_L^1,H_L^2, … } ]$,假设这些随机变量相互独立且服从相同分布,分布函数用${F _{\rm h}}$表示. 由于该分布未知,故针对该序列研究其阈值$u$的超出量的分布,如式(14)所示,其中$y$为超出量. 需要注意的是,阈值$u$是极值理论[32]的一个研究对象,它不同于待求解的阈值${\delta _{\rm h}}$. 由于原始的分布函数${F _{\rm h}}$未知,假设阈值$u$服从广义的Pareto分布,分布函数${F_u}$如式(15)所示. 其中,$\xi $为形状参数,$\mu $为位置参数(实际取$\mu=0 $),$\tilde \sigma = \sigma + \xi \left( {u - \mu } \right)$为尺度参数. 文献[33]给出了参数$\hat \xi $和$\hat \sigma $的估计方法,进而可根据式(16)计算出决策值${z_{{q _{\rm h}}}}$. 其中,$u$的值根据经验设置为0.98倍样本最大值,${N^*}$为观测样本的总数量,${N_u}$为超过观测值超过阈值$u$的样本数量. 对于一个给定的危险系数$ {q _{\rm h}} $,如${q _{\rm h}} = 1E - 5$,令${\delta _{\rm h}} = {z_{{q _{\rm h}}}}$即可得到相应的阈值. 作为熵值序列的极值,${\delta _{\rm h}}$满足式(12).
$ Pr \left\{ {{H_L} < u + y|{H_L} > u} \right\} = \frac{{{F _{\rm h}}\left( {u + y} \right) - {F _{\rm h}}\left( u \right)}}{{1 - {F _{\rm h}}\left( u \right)}},\;\;y > 0 \text{,} $
$ {F_u} = 1 - {\left( {1 + \frac{{\xi y}}{{\tilde \sigma }}} \right)^{ - \tfrac{1}{\xi }}},\;\;y > 0,\;\;1 + \frac{{\xi y}}{{\tilde \sigma }} > 0 \text{,} $
$ {z_{{q _{\rm h}}}} = u + \frac{{\hat \sigma }}{{\hat \xi }}\left( {{{\left( {\frac{{{N^*}\left( {1 - {q _{\rm h}}} \right)}}{{{N_u}}}} \right)}^{ - \hat \xi }} - 1} \right). $
接着,确定阈值${\delta _{\rm k}}$. 记$K$值随机变量构成的序列为$ [ {K_l^1,K_l^2, … } ]$,假设这些随机变量相互独立且服从相同分布,分布函数用${F _{\rm k}}$表示. 由式(11)可知,在一个统计周期内,轨迹序列中连续1的数量为首次出现0之前1出现的次数. 这类似于抛硬币实验中,硬币首次出现反面时正面已经出现的次数. 因此,利用几何分布拟合随机变量$K$的分布,如式(17)所示. 令${p_l}$表示虚拟会话选择链路$l$的概率,基于采集的数据集,期望$E\left( {{K_l}} \right)$可以通过统计分析得到. 进一步,根据一阶矩估计$E\left( {{K_l}} \right) = {1 / { {{\hat p_l}}}}$,可计算得到$ {{\hat p_l}}$和$ {{\hat F _{\rm k}}}$,进而可计算出决策值${z_{{q _{\rm k}}}}$,如式(18)所示. 对于一个给定的危险系数${q _{\rm k}}$,如${q _{\rm k}} = 1E- 5$,令${\delta _{\rm k}} = {z_{{q _{\rm k}}}} + 1$即可得相应的阈值. 作为分布函数${F _{\rm k}}$的边界,${\delta _{\rm k}}$满足式(13).
$ {{\hat F _{\rm k}}} = \sum\limits_{i \in \mathbb{N}} {{{\left( {1 - \hat p} \right)}^{i - 1}} \times } \hat p \text{,} $
$ {z_{{q _{\rm k}}}} = \mathop {\min }\limits_{ {{\hat F _{\rm k}}} \geqslant 1 - {q _{\rm k}}} K. $
需要说明的是,式(16)与式(18)中阈值设定依赖于参数$u$或$q$,但二者对模型性能的影响均较为有限. 具体地,$u$表示尾部数据的截断阈值,尽管其取值会影响参数$\sigma $与$\xi $的估计,但在典型取值范围内(如样本最大值的$0.90\sim 0.99$倍),模型性能基本保持稳定,表明其对参数$u$的选择具有一定鲁棒性. 参数$q$用于调节异常判别的严格程度,如图4(a)所示,当$q$取值范围是$1E - 5\sim 1E - 2$时,模型检测性能随着$q$的减小逐渐趋于稳定,表明模型对$q$的设定具有良好的适应性.
本节首先介绍实验环境以及采集的数据集,接着通过实验验证威胁模型中的观点以及CLV方法的可行性、有效性、可扩展性和鲁棒性.
实验环境运行于Ubuntu 20.04系统的虚拟主机,物理主机的CPU为Intel Core i7-11800H,核心数为8,内存32 GB. 实验使用Mininet[34]模拟智能体网络,使用交换机模拟Agent worker,如图5所示. 在该网络中,正常链路的带宽设置为100 Mbps,伪造链路的带宽设置为50 Mbps,表明伪造链路和恶意Agent worker的引入会降低消息的传输和处理性能. 此外,分别基于Scapy[35]和PyPcap[36]实现不同性能的中继模块,基于iPerf[37]工具生成背景流量. 如表2所示,实验考虑了6种攻击场景,包括高或低的传输或转发时延、是否过载Agent协调器,以及是否存在背景流量.
关于不同攻击场景的说明如下:Case 1表示高传输时延,且无扰动Agent协调器和背景流量的情况. Case 2表示高转发时延,且无扰动和背景流量的情况. Case 3表示低传输和转发时延,无扰动但背景流量致使伪造链路过载的情况. Case 4表示低传输和转发时延,且扰动和背景流量的情况. Case 5表示低传输和转发时延,无背景流量但存在扰动的情况. Case 6表示低传输和转发时延,且存在扰动和背景流量的情况. 特别地,Case 4、Case 5、Case 6对应低时延网络环境下的攻击场景.
关于参数设置的说明如下:在Case 1中,传输时延设置为5 ms,与文献[14,16-17], 保持一致. 其他情况下传输时延设置均为0.5 ms,以指示低时延的网络环境. 在Case 2中,相关方法[14,16,20]基于Scapy实现中继模块,其相比PyPcap实现的中继模块具有更高的转发时延. 在Case 3中,背景流量设置为35~45 Mbps,使得链路过载但不会导致中继模块崩溃. 同时,仅采集链路过载时的时延用于统计分析. 在Case 5中,设置攻击速率为每秒1000报文,以扰动Agent协调器但避免其服务不可用. 在Case 6中,设置背景流量为20~30 Mbps,以避免链路过载.
表3给出了实验数据集的描述. 其中,特征向量为$( {{d_{\rm cs}},{d_{\rm sc}},{d_l},load} )$,${d_{\rm cs}}$和${d_{\rm sc}}$分别表示Agent协调器到Agent worker以及Agent worker到Agent协调器之间的测量时延,${d_l}$和$load$分别表示链路$l$的测量时延和负载. 相较于文献[16],此处增加了链路负载的特征,用以计算链路质量. 采样周期数为链路探测的次数,实际每隔2 s探测一次网络中的所有链路.
该实验旨在验证威胁模型中提出的观点:1)在Case 1~3的攻击场景下,当链路传输时延较高,中继模块性能较低(亦即转发时延较高)或伪造链路过载时,伪造链路与正常链路之间的性能差异是显著的. 2)在Case 4~6的攻击场景下,当攻击违背Case 1~3情况时,伪造链路与正常链路之间的性能差异不再明显,进而导致现有方法的检测性能不佳.
实验结果如图6所示. 在Case 1~3的攻击场景下,正常链路与伪造链路上测量时延的分布差异是明显的,相关工作[14,16,20]表明,这些方法均能取得理想的检测效果. 然而,在Case 4~6的攻击场景下,正常链路与伪造链路上测量时延的分布差异不再明显. 根据4.4节中的实验结果可知,此时相关方法均表现出较低的检测率和较高的误判率. 因此,威胁模型中的观点得到了验证.
为了说明低时延网络中CLV方法的可行性,需要验证2点:1)对于关联链路而言,伪造链路的存在是否会导致它与正常链路之间产生可观测的性能差异. 2)在低时延网络中,本文使用的统计特征能否指示关联链路中的伪造链路.
首先,对第1点进行验证. 为展示链路之间的性能差异,采集多种攻击场景下100个采样周期的数据,计算每个周期关联链路的链路质量,如图7所示. 可见,在所有场景下,正常关联链路之间的性能差异非常细微,而由正常链路和伪造链路构成的关联链路之间的性能差异相对明显且容易观测到. 因此,第1点得到了验证.
接着,对第2点进行验证. 考虑低时延的网络环境,仅关注Case 4~6的攻击场景. 同时,获取这些攻击场景下100个采样周期的数据,利用多径传输模拟方法得到链路的轨迹序列,并计算统计特征熵值和$K$值. 如图8所示,对于正常关联链路,轨迹序列中0-1分布的香农熵值在各个时刻均接近1,表明正常关联链路之间的性能差异非常细微. 而对于包含一条伪造链路的异常关联链路,轨迹序列中0-1分布的香农熵值明显低于1,表明关联链路中正常链路与伪造链路之间存在可观测的性能差异. 如图9所示,对于正常关联链路和异常关联链路,其轨迹序列中连续1的数量分布不同,且前者$K$值一般大于后者. 综上,这2个统计特征能够用于指示伪造链路的存在,故第2点得到了验证. 需要说明的是,图9中二者的熵值序列长度不同,这是因为它们的$K$值不同,导致统计周期的大小不同,如式(11)所示.
为了验证低时延网络中CLV方法的有效性,获取Case 4~6攻击场景下采集的数据集,对比CLV方法与相关工作中方法的检测效果,评估指标包括检测率TPR、误判率FPR、检测准确率Pr和F1分数. 另外,相关工作中检测方法的具体实现在威胁模型部分已经进行了介绍,此处不予赘述. 需要说明的是,该实验严格按照相关工作的描述进行参数设置和模型调优. 对于RLV[16],使用sklearn训练其使用的机器学习模型,并采用网格搜索寻找最优的超参数. 对于TopoGuard+[17],无需设置超参数. 对于TrustTopo[20],设置超参数$c = 0.5$使得模型的检测效果最佳.
表4~6给出了不同攻击场景下相关方法的检测结果,从中可以发现,相关方法无法兼顾检测率和误判率. RLV相较于TrustTopo和TopoGuard+具有更高的检测率和误判率. 其原因在于,这些方法直接基于测量的时延训练检测模型,难以得到能够准确识别伪造链路的阈值. 过高的阈值导致较低的检测率和误判率,而过低的阈值导致较高的检测率和误判率. CLV-OPT和CLV-RAW分别表示是否采用提出的数据处理方法. 经过转换的统计特征具有更清晰的界限,并且阈值的确定过程具有良好的可解释性,故针对该数据集,CLV方法能够取得超过99%的检测率和约0%的误判率. 由于该数据集的规模较小,CLV-OPT与CLV-RAW的性能差异并不明显,在扩展性和鲁棒性验证的实验中,二者的差异将会凸显出来.
为了验证低时延网络中CLV方法的可扩展性,仅考虑包含扰动和背景流量的Case 6场景. 以交换机模拟Agent worker,设置数量10~100. 同时,以$AUC$值为指标设置CLV方法中超参数的值. $AUC$是评估分类器性能的常用指标,其值为0~1,且越接近1表明分类器的性能越好. 针对CLV方法,超参数取值对检测性能的影响如图4所示.
图4(a)显示了危险系数的影响,它对应式(16)中的参数${q _{\rm h}}$和式(18)中的参数${q _{\rm k}}$. 尽管当危险系数低于$ 1\mathrm{E}-3 $时检测性能保持不变,但为了取得尽可能低的误判率,设置危险系数值为$ 1\mathrm{E}-5 $. 图4(b)和图4(c)分别显示了折扣率和有效时延上界的影响,它们对应于式(10)中的参数${\gamma _0}$和式(8)中的参数${d_{\rm lim}}$. 显然,最佳的折扣率值为0.9,最佳的有效时延上界值为5. 图4(d) 表明了统计周期次数的影响. 根据${T_{\rm det}} = Interval \times {T_{\rm sta}}$知,统计次数越少,检测时延越小. 而统计次数超过10时检测性能保持不变,故将其设置为10.
不同网络规模下相关方法的检测性能如图10所示. 可以发现,随着Agent worker数量的增加,这些方法的检测性能总体呈现出不同程度的下降. 其中,TopoGuard+方法的检测性能变化较小,始终保持在较低的状态,这归因于其过高的检测阈值. 然而,这也会导致过低的检测率. 因此,它们难以适用于低时延网络. 相反,RLV和TrustTopo方法具有较高的检测率和误判率,这归因于其过低的检测阈值. 由于误判率过高,如在大部分情况下超过30%,它们同样难以适用于低时延网络. 此外,CLV-RAW相较于CLV-OPT性能下降得更快,这表明数据处理方法能够有效缓解测量误差,提高模型的可扩展性. 相较于这些方法,CLV方法表现出更好的可扩展性和稳定性. 因此,CLV方法的可扩展性得到了验证.
在该文中,鲁棒性特指检测方法对背景流量和扰动的容忍能力. 为了验证CLV方法的鲁棒性,仅关注包含攻击和背景流量的Case 6场景,并引入不同数量的伪造链路,以及不同大小的链路传输时延. 需要说明的是,为了确保一定强度的背景流量顺利生成,限制Agent worker的数量为32~48,伪造链路的数量为1~9. 当伪造链路的数量为9时,受影响的Agent worker数量为18,这意味着网络中约一半的Agent worker会受到攻击的影响. 这种极端的场景有利于评估CLV方法的鲁棒性.
如图11和图12所示,随着伪造链路数量的增加,CLV-RAW和CLV-OPT方法的检测性能均会有所下降. 其原因在于,随着伪造链路数量的增加,网络波动更加剧烈,导致采集到的无效数据增多,进而影响模型的检测性能. 尽管如此,当伪造链路的数量达到9时,CLV-RAW仍能够取得约95.61%的检测率和约3.7%的误判率,CLV-OPT方法可取得约96.66%的检测率和约2.36%的误判率. 实验结果表明,CLV方法具有较强的抗扰动能力.
如图13和图14所示,链路的传输时延会影响CLV方法的检测性能. 具体地,链路传输时延越小,伪造链路与正常链路的性能差异越小,CLV方法的检测性能相对越差. 当链路传输时延为0.1 ms时,CLV-RAW方法仅能取得4.97%的检测率,而CLV-OPT依然能够取得约95.23%的检测率和约0.28%的误判率. 该实验结果表明,所提出的数据处理方法能够有效缓解测量误差的影响,进而增强模型的鲁棒性. 综上,CLV方法的鲁棒性得到了验证.
本文针对低时延智能体网络中的伪造链路攻击检测问题,提出了一种基于关联链路验证的检测方法. 该方法通过引入数据预处理机制提升测量鲁棒性,利用多径传输模拟将关联链路间的性能差异转化为统计特征,并结合极值理论与概率分布拟合确定检测阈值,从而实现对伪造链路的有效识别. 为验证方法的有效性,本文从可行性、检测性能、可扩展性、鲁棒性等方面开展实验,结果表明CLV在不同网络规模和攻击环境下均具有良好的适应性和稳定性.
作者贡献声明:张峰、俸皓、刘玉明提出了智能体网络伪造链路攻击检测问题,设计了整体方案并撰写论文;周秀民、章茂淳、陈镇铭负责部分算法和实验.

参考文献 引证文献
排序方式:
1
Liu Yan, Guo Bin, Li Nuo, et al. CrowdTransfer: Enabling crowd knowledge transfer in AIoT community[J]. IEEE Communications Surveys & Tutorials, 2024, 27(2): 1191−1237
2
Liu Sicong, Guo Bin, Fang Cheng, et al. Enabling resource-efficient AIoT system with cross-level optimization: A survey[J]. IEEE Communications Surveys & Tutorials, 2023, 26(1): 389−427
3
Chen Weize, You Ziming, Li Ran, et al. Internet of agents: Weaving a Web of heterogeneous agents for collaborative intelligence[C]//Proc of the Int Conf on Learning Representations (ICLR). San Diego, CA: OpenReview. net, 2025: 1−32
4
Agent Network Protocol Contributors. Agent Network Protocol (ANP). [EB/OL]. [2025-7-16]. https://github.com/Agent-network-protocol/AgentNetworkProtocol
5
Fu Bo, Smith W, Rizzo D, et al. Robust task scheduling for heterogeneous robot teams under capability uncertainty[J]. IEEE Transactions on Robotics, 2022, 39(2): 1087−1105
6
Soltani S, Amanlou A, Shojafar M, et al. Security of topology discovery service in SDN: Vulnerabilities and countermeasures[J]. IEEE Open Journal of the Communications Society, 2024, 5: 3410−3450
7
郑正, 徐明伟, 李琦, 等. 网络拓扑污染攻击防御机制研究[J]. 计算机研究与发展, 2018, 55(1): 207−216
Zheng Zheng, Xu Mingwei, Li Qi, et al. Defending against SDN network topology poisoning attacks[J]. Journal of Computer Research and Development, 2018, 55(1): 207−216 (in Chinese)
8
Taheri M, Khorasani K, Shames I, et al. Undetectable cyber attacks on communication links in multi-agent cyber-physical systems[C]//Proc of the 59th IEEE Conf on Decision and Control (CDC). Piscataway, NJ: IEEE, 2020: 3764−3771
9
Ding Lifu, Yan Gangfeng. A survey of the security issues and defense mechanisms of multi-agent systems[J]. CAAI Transactions on Intelligent Systems, 2020, 15(3): 425−434
10
丁礼富, 严刚峰. 多智能体系统安全性问题及防御机制综述[J]. 信息与控制, 2018, 47(6): 651−660
Ding Lifu, Yan Gangfeng. A survey of security issues and defense mechanisms in multi-agent systems[J]. Information and Control, 2018, 47(6): 651−660 (in Chinses
11
Cao Yinzhi, Yu A F, Aday A, et al. Efficient repair of polluted machine learning systems via causal unlearning[C]//Proc of the 2018 on Asia Conf on Computer and Communications Security (ASIACSS). New York: ACM, 2018: 735−747
12
Zeng Zengri, Peng Wei, Zeng Detian. Improving the stability of intrusion detection with causal deep learning[J]. IEEE Transactions on Network and Service Management, 2022, 19(4): 4750−4763
13
Smyth D, McSweeney S, O’Shea D, et al. Detecting link fabrication attacks in software-defined networks[C]//Proc of the 26th Int Conf on Computer Communication and Networks (ICCCN). Piscataway, NJ: IEEE, 2017: 1(1): 1−8
14
Al Salti I, Zhang N. LINK-GUARD: An effective and scalable security framework for link discovery in SDN networks[J]. IEEE Access, 2022, 10(10): 130233−130252
15
Soltani S, Shojafar M, Mostafaei H, et al. Link latency attack in software-defined networks[C]//Proc of the 17th Int Conf on Network and Service Management (CNSM). Piscataway, NJ: IEEE, 2021, 1(1): 187−193
16
Soltani S, Shojafar M, Mostafaei H, et al. Real-time link verification in software-defined networks[J]. IEEE Transactions on Network and Service Management, 2023, 20(3): 3596−3611
17
Skowyra R, Xu Lei, Gu Guofei, et al. Effective topology tampering attacks and defenses in software-defined networks[C]//Proc of the 48th Annual IEEE/IFIP Int Conf on Dependable Systems and Networks (DSN). Piscataway, NJ: IEEE, 2018: 374−385
18
Marin E, Bucciol N, Conti M. An in-depth look into SDN topology discovery mechanisms: Novel attacks and practical countermeasures[C]//Proc of the 2019 ACM SIGSAC Conf on Computer and Communications Security, New York, ACM, 2019: 1101−1114
19
Chou L D, Liu C C, Lai M S, et al. Behavior anomaly detection in SDN control plane: A case study of topology discovery attacks[J]. Wireless Communications and Mobile Computing, 2020, 2020(1): 8898949
20
Huang Xinli, Shi Peng, Liu Yufei, et al. Towards trusted and efficient SDN topology discovery: A lightweight topology verification scheme[J]. Computer Networks, 2020, 170: 107119
21
Shrivastava P, Agarwal A, Kataoka K. Detection of topology poisoning by silent relay attacker in SDN[C]//Proc of the 24th annual Int Conf on Mobile Computing and Networking. New York: ACM, 2018: 792−794
22
Wang Jiadai, Liu Jiajia, Guo Hongzhi, et al. Deep reinforcement learning for securing software defined industrial networks with distributed control plane[J]. IEEE Transactions on Industrial Informatics, 2022, 18(6): 4275−4285
23
Kong Dezhang, Shen Yi, Chen Xiang, et al. Combination attacks and defenses on SDN topology discovery[J]. IEEE/ACM Transactions on Networking, 2023, 31(2): 904−919
24
Azzouni A, Boutaba R, Trang N T M, et al. sOFTDP: Secure and efficient OpenFlow topology discovery protocol[C]//Proc of 2018 IEEE/IFIP Network Operations and Management Symp. Piscataway, NJ: IEEE, 2018: 1−7
25
Nehra A, Tripathi M, Gaur M S, et al. SLDP: A secure and lightweight link discovery protocol for software defined networking[J]. Computer Networks, 2019, 150: 102−116
26
Jia Yongze, Xu Lei, Yang Yuwang, et al. Lightweight automatic discovery protocol for OpenFlow-based software defined networking[J]. IEEE Communications Letters, 2020, 24(2): 312−315
27
Gu Yongpu, Li Dong, Yu Junqing. IM-OFDP: An improved openflow-based topology discovery protocol for software defined network[C]//Proc of 2020 IFIP Networking Conf (Networking). Piscataway, NJ: IEEE, 2020: 628−630
28
Hauser F, Schmidt M, Haberle M, et al. P4-MACsec: Dynamic topology monitoring and data layer protection with MACsec in P4-based SDN[J]. IEEE Access, 2020, 8: 58845−58858
29
Jero S, Koch W, Skowyra R, et al. Identifier binding attacks and defenses in software-defined networks[C]//Proce of the 26th USENIX Conf on Security Symp. Berkeley, CA: USENIX Association, 2017: 415−432
30
Zhou Qizhao, Yu Junqing, Li Dong. A dynamic and lightweight framework to secure source addresses in the SDN-based networks[J]. Computer Networks, 2021, 193: 108075
31
Meena R C, Bhatia S, Jhaveri R H, et al. HyPASS: Design of hybrid-SDN prevention of attacks of source spoofing with host discovery and address validation[J]. Physical Communication, 2022, 55: 101902
32
Haan L, Ferreira A. Extreme Value Theory: An Introduction[M]. New York: Springer, 2006: 65−126
33
Siffer A, Fouque P A, Termier A, et al. Anomaly detection in streams with extreme value theory[C]//Proc of the 23rd ACM SIGKDD Int Conf on Knowledge Discovery and Data Mining (KDD). New York: ACM, 2017: 1067−1075
34
Mininet Project Contributors. An instant virtual network on your laptop (or other PC)[EB/OL]. [2025-07-16]. https://mininet.org/
35
Scapy community. Scapy: The Python-based interactive packet manipulation program &amp; library[EB/OL]. [2025-7-16]. https://github.com/secdev/scapy
36
Arturo Filastò. PyPcap - Python libpcap module. [EB/OL]. [2025-07-16]. https://github.com /pynetwork/pypcap
37
ESnet/Lawrence Berkeley National Laboratory. iPerf-The ultimate speed test tool for TCP UDP and SCTP[EB/OL]. [2025-07-16]. https://iperf.fr
2025年第62卷第10期
134
62
0
引用本文
BibTeX
文章信息
doi: 10.7544/issn1000-1239.202550399
  • 接收时间:2025-06-04
  • 首发时间:2026-11-10
  • 出版时间:2025-10-01
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2025-06-04
  • 修回日期:2025-07-16
基金
作者信息
    桂林电子科技大学计算机与信息安全学院 广西桂林 541004

通讯作者:

刘玉明()
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN/10.7544/issn1000-1239.202550399
分享至
全文二维码

扫描看全文

引用本文
BibTeX
本文的引用情况
2种不同金属材料的力学参数
科
Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)
属
Genus
种数
Number of
species
占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏