sibsutisВестник СибГУТИThe Herald of the Siberian State University of Telecommunications and Information Science1998-6920СибГУТИ10.55648/1998-6920-2026-20-2-28-44sibsutis-1068Research ArticleСтатьиАнализ устойчивости модели адаптивной аутентификации к целевым оптимизационным атакам имитации гибридного цифрового отпечаткаAnalysis of the resilience of the adaptive authentication model to adversarial attacks via hybrid digital fingerprint imitationhttps://orcid.org/0000-0002-1143-5275СаломатинАлександр АлександровичSalomatinAlexander Alexandrovich

к.т.н., старший научный сотрудник, Институт проблем управления им. В.А. Трапезникова РАН

sandr@ipu.ruhttps://orcid.org/0000-0002-8049-851XШироковАлександр СергеевичShirokovAlexander Sergeevich

научный сотрудник, Институт проблем управления им. В.А. Трапезникова РАН (ИПУ РАН)

shiras@ipu.ruМельниковАндрей КимовичMelnikovAndrey Kimovich

к.т.н., доцент ВАК, главный научный сотрудник, АО «Вычислительные решения»

ak@comp-sol.ruИнститут проблем управления им. В.А. Трапезникова РАН (ИПУ РАН, 117342, г. Москва, ул. Профсоюзная, д. 65, стр. 2)РоссияV.A. Trapeznikov Institute of Control Sciences of the Russian Academy of Sciences (ICP RAS, 117342, Moscow, Profsoyuznaya St., 65, building 2)Russian FederationАО «Вычислительные решения» (117587, город Москва, Варшавское ш, д. 125)РоссияJSC "Computational Solutions" (117587, Moscow, Varshavskoe shosse, no. 125)Russian Federation202622062026202Вестник СибГУТИ2844Copyright © Саломатин А.А., Широков А.С., Мельников А.К., 20262026Саломатин А.А., Широков А.С., Мельников А.К.Salomatin A.A., Shirokov A.S., Melnikov A.K.Данная работа распространяется под лицензией Creative Commons Attribution 4.0.This work is licensed under a Creative Commons Attribution 4.0 License.https://vestnik.sibsutis.ru/jour/article/view/1068

В статье исследуется устойчивость модели адаптивной аутентификации к целевым оптимизационным атакам имитации гибридного цифрового отпечатка. Предложено адверсариальное расширение модели адаптивной аутентификации, основанной на цифровом отпечатке, объединяющем технические и поведенческие атрибуты. Атака формализована как задача условной оптимизации с раздельными величинами допустимых возмущений технических (εd) и поведенческих (εb) атрибутов в условиях ограниченной обратной связи. Для оценки устойчивости модели введена вероятностная мера устойчивости R, согласованная с классическими метриками FAR/FRR. Экспериментальная проверка проведена на датасете из 1280 сессий (32 пользователя, 30 признаков) с использованием алгоритма CMA-ES. Результаты экспериментов показали, что модели с адаптивным взвешиванием обеспечивают двукратное повышение устойчивости относительно модели с равными весами в зоне реалистичных атак (εb≤0.2): R=0.66-0.79 против 0.39–0.53, и снижают вероятность успешной атаки до 0.36 при стратегии оптимизации технических и поведенческих признаков. Экспериментально установлено, что автоматическое снижение весов при росте дисперсии признаков затрудняет целевые оптимизационные атаки злоумышленников. Выявленная уязвимость защиты при εd≥0.5 (R≤0.07) определяет границы применимости модели и обосновывает необходимость многофакторной аутентификации при высокой доле возмущений.

The article investigates the resilience of the adaptive authentication model to adversarial attacks via hybrid digital fingerprint imitation. An adversarial extension of the adaptive authentication model based on a digital fingerprint combining technical and behavioral attributes is proposed. The attack is formalized as a conditional optimization problem with separate values of permissible perturbations of technical (εd) and behavioral (εb) attributes under conditions of limited feedback. To evaluate model resilence, a probabilistic resilience measure R is introduced, consistent with the classical FAR/FRR metrics. Experimental validation was performed on a dataset of 1280 sessions (32 users, 30 attributes) using the CMA-ES algorithm. The experimental results showed that models with adaptive weighting provide a twofold increase in resilience compared to the model with equal weights in the realistic attack regime (εd ≤0.2): R=0.66-0.79 versus 0.39–0.53, and reduce the probability of a successful attack to 0.36 under a strategy for optimizing technical and behavioral features. It has been experimentally established that the automatic reduction of weights with an increase in the variance of features complicates adversarial attacks by attackers. The identified security vulnerability at εd ≥0.5 (R≤0.07) defines the limits of the model's applicability and justifies the need for multifactor authentication with a high proportion of perturbations.

гибридный цифровой отпечатокадаптивная аутентификацияцелевые оптимизационные атакиустойчивость аутентификацииповеденческая биометриякибербезопасностьhybrid digital fingerprintadaptive authenticationadversarial attacksauthentication resiliencebehavioral biometricscybersecurityReferencesВехов В.Б., Смушкин А.Б. Криминалистическое исследование цифровых отпечатков компьютерных устройств // Всероссийский криминологический журнал. – 2024. – Т. 18, № 4. – С. 390-397. – DOI 10.17150/2500-4255.2024.18(4).390-397. – EDN PCRWRCVekhov V.B., Smushkin A.B. Kriminalisticheskoe issledovanie tsifrovykh otpechatkov komp'yuternykh ustroistv // Vserossiiskii kriminologicheskii zhurnal. – 2024. – T. 18, № 4. – S. 390-397. – DOI 10.17150/2500-4255.2024.18(4).390-397. – EDN PCRWRCСаломатин А.А., Лукинова О.В. Теоретические основы адаптации легитимизации пользователя на основе динамики вектора атрибутов цифрового отпечатка // Известия Юго-западного государственного университета. – 2026. – (в печати).Salomatin A.A., Lukinova O.V. Teoreticheskie osnovy adaptatsii legitimizatsii pol'-zovatelya na os-nove dinamiki vektora atributov tsifrovogo otpechatka // Izvestiya Yugo-zapadnogo gosudarstvennogo universiteta. – 2026. – (in press).Bonneau J., Herley C., van Oorschot P. C., Stajano F. Passwords and the Evolution of Imperfect Authentication // Communications of the ACM. – 2015. – Vol. 58, No. 7. – P. 78–87.Bonneau J., Herley C., van Oorschot P. C., Stajano F. Passwords and the Evolution of Imperfect Au-thentication // Communications of the ACM. – 2015. – Vol. 58, No. 7. – P. 78–87.Bonneau J., Herley C., van Oorschot P. C., Stajano F. The Quest to Replace Passwords: A Framework for Comparative Evaluation of Web Authentication Schemes // IEEE Symposium on Security and Privacy. – San Francisco, CA, USA. – 2012. – P. 553–567. – DOI: 10.1109/SP.2012.44Bonneau J., Herley C., van Oorschot P. C., Stajano F. The Quest to Replace Passwords: A Framework for Comparative Evaluation of Web Authentication Schemes // IEEE Symposium on Security and Pri-vacy. – San Francisco, CA, USA. – 2012. – P. 553–567. – DOI: 10.1109/SP.2012.44Das A., Bonneau J., Caesar M., Borisov N., Wang X. The Tangled Web of Password Reuse // Proceedings of the Network and Distributed System Security Symposium. – 2014. – DOI: 10.14722/ndss.2014.23357Das A., Bonneau J., Caesar M., Borisov N., Wang X. The Tangled Web of Password Reuse // Proceed-ings of the Network and Distributed System Security Symposium. – 2014. – DOI: 10.14722/ndss.2014.23357Juels A., Rivest R. Honeywords: Making Password-Cracking Detectable // Proceedings of the 2013 ACM SIGSAC Conference on Computer and Communications Security. – 2013. – P. 145–160. – DOI: 10.1145/2508859.2516671Juels A., Rivest R. Honeywords: Making Password-Cracking Detectable // Proceedings of the 2013 ACM SIGSAC Conference on Computer and Communications Security. – 2013. – P. 145–160. – DOI: 10.1145/2508859.2516671Eberz S., Rasmussen K. B., Lenders V., Martinovic I. Evaluating Behavioral Biometrics for Continuous Authentication: Challenges and Metrics // Proceedings of the 2017 ACM on Asia Conference on Computer and Communications Security. – 2017. – P. 386–399. – DOI: 10.1145/3052973.3053032Eberz S., Rasmussen K. B., Lenders V., Martinovic I. Evaluating Behavioral Biometrics for Continu-ous Authentication: Challenges and Metrics // Proceedings of the 2017 ACM on Asia Conference on Computer and Communications Security. – 2017. – P. 386–399. – DOI: 10.1145/3052973.3053032Viswanath B., Bashir M. A., Crovella M., Guha S., Gummadi K. P., Krishnamurthy B., Mislove A. Towards Detecting Anomalous User Behavior in Online Social Networks // USENIX Security Symposium. – 2014. – P. 223–238.Viswanath B., Bashir M. A., Crovella M., Guha S., Gummadi K. P., Krishnamurthy B., Mislove A. Towards Detecting Anomalous User Behavior in Online Social Networks // USENIX Security Sympo-sium. – 2014. – P. 223–238.Monrose F., Rubin A. Authentication via Keystroke Dynamics // Proceedings of the 4th ACM Conference on Computer and Communications Security. – 1997. – P. 48–56. – DOI: 10.1145/266420.266434Monrose F., Rubin A. Authentication via Keystroke Dynamics // Proceedings of the 4th ACM Confer-ence on Computer and Communications Security. – 1997. – P. 48–56. – DOI: 10.1145/266420.266434Killourhy K., Maxion R. Comparing Anomaly Detection Algorithms for Keystroke Dynamics // IEEE/IFIP International Conference on Dependable Systems and Networks. – Lisbon, Portugal. – 2009. – P. 125–134. – DOI: 10.1109/DSN.2009.5270346.Killourhy K., Maxion R. Comparing Anomaly Detection Algorithms for Keystroke Dynamics // IEEE/IFIP International Conference on Dependable Systems and Networks. – Lisbon, Portugal. – 2009. – P. 125–134. – DOI: 10.1109/DSN.2009.5270346.Shen C., Cai Z., Guan X., Maxion R. Performance Evaluation of Anomaly-Detection Algorithms for Mouse Dynamics // Computers & Security. – 2014. – Vol. 45. – P. 156–171. – DOI: 10.1016/j.cose.2014.05.002Shen C., Cai Z., Guan X., Maxion R. Performance Evaluation of Anomaly-Detection Algorithms for Mouse Dynamics // Computers & Security. – 2014. – Vol. 45. – P. 156–171. – DOI: 10.1016/j.cose.2014.05.002Serwadda A., Phoha V. When Kids’ Toys Breach Mobile Phone Security // Proceedings of the 2013 ACM SIGSAC Conference on Computer and Communications Security. – 2013. – P. 599–610. – DOI: 10.1145/2508859.2516659Serwadda A., Phoha V. When Kids’ Toys Breach Mobile Phone Security // Proceedings of the 2013 ACM SIGSAC Conference on Computer and Communications Security. – 2013. – P. 599–610. – DOI: 10.1145/2508859.2516659Biggio B., Roli F. Wild Patterns: Ten Years after the Rise of Adversarial Machine Learning // Pattern Recognition. – 2018. – Vol. 84. – P. 317–331. – DOI: 10.1016/j.patcog.2018.07.023.Biggio B., Roli F. Wild Patterns: Ten Years after the Rise of Adversarial Machine Learning // Pattern Recognition. – 2018. – Vol. 84. – P. 317–331. – DOI: 10.1016/j.patcog.2018.07.023.Goodfellow I. J., Shlens J., Szegedy C. Explaining and Harnessing Adversarial Examples // arXiv. – 2015. – DOI: 10.48550/arXiv.1412.6572Goodfellow I. J., Shlens J., Szegedy C. Explaining and Harnessing Adversarial Examples // arXiv. – 2015. – DOI: 10.48550/arXiv.1412.6572Papernot N. et al. Practical Black-Box Attacks Against Machine Learning // Proceedings of the 2017 ACM on Asia Conference on Computer and Communications Security. – 2017. – P. 506–519. – DOI: 10.1145/3052973.3053009Papernot N. et al. Practical Black-Box Attacks Against Machine Learning // Proceedings of the 2017 ACM on Asia Conference on Computer and Communications Security. – 2017. – P. 506–519. – DOI: 10.1145/3052973.3053009Sommer R., Paxson V. Outside the Closed World: On Using Machine Learning for Network Intrusion Detection // IEEE Symposium on Security and Privacy. – Oakland, CA, USA. – 2010. – P. 305–316. – DOI: 10.1109/SP.2010.25.Sommer R., Paxson V. Outside the Closed World: On Using Machine Learning for Network Intrusion Detection // IEEE Symposium on Security and Privacy. – Oakland, CA, USA. – 2010. – P. 305–316. – DOI: 10.1109/SP.2010.25.Dalvi N. et al. Adversarial Classification // Proceedings of the Tenth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. – 2004. – P. 99–108. – DOI: 10.1145/1014052.1014066Dalvi N. et al. Adversarial Classification // Proceedings of the Tenth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. – 2004. – P. 99–108. – DOI: 10.1145/1014052.1014066Li B., Vorobeychik Y. Scalable Optimization of Randomized Operational Decisions in Adversarial Classification Settings // Proceedings of the Eighteenth International Conference on Artificial Intelligence and Statistics. – 2015. – Vol. 38. – P. 599–607.Li B., Vorobeychik Y. Scalable Optimization of Randomized Operational Decisions in Adversarial Classification Settings // Proceedings of the Eighteenth International Conference on Artificial Intelli-gence and Statistics. – 2015. – Vol. 38. – P. 599–607.Barreno M. et al. The Security of Machine Learning // Machine Learning. – 2010. – Vol. 81. – P. 121–148. – DOI: 10.1007/s10994-010-5188-5Barreno M. et al. The Security of Machine Learning // Machine Learning. – 2010. – Vol. 81. – P. 121–148. – DOI: 10.1007/s10994-010-5188-5Chandola V. et al. Anomaly Detection: A Survey // ACM Computing Surveys. – 2009. – Vol. 41, No. 3. – Article 15. – P. 1–58. – DOI: 10.1145/1541880.1541882Chandola V. et al. Anomaly Detection: A Survey // ACM Computing Surveys. – 2009. – Vol. 41, No. 3. – Article 15. – P. 1–58. – DOI: 10.1145/1541880.1541882Apify. fingerprint-generator: Realistic browser fingerprint generator [Электронный ресурс]. – 2024. – URL: https://github.com/apify/fingerprint-generator (дата обращения: 06.04.2026).Apify. fingerprint-generator: Realistic browser fingerprint generator. – 2024. – URL: https://github.com/apify/fingerprint-generator (accessed: 06.04.2026).Andriamilanto N., Allard T., Le Guelvouit G., Garel A. A Large-scale Empirical Analysis of Browser Fingerprints Properties for Web Authentication // ACM Transactions on the Web. – 2021. – Vol. 16, No. 1. – Art. 4. – DOI: 10.1145/3478026Andriamilanto N., Allard T., Le Guelvouit G., Garel A. A Large-scale Empirical Analysis of Browser Fingerprints Properties for Web Authentication // ACM Transactions on the Web. – 2021. – Vol. 16, No. 1. – Art. 4. – DOI: 10.1145/3478026Barford P., Crovella M. Generating Representative Web Workloads for Network and Server Performance Evaluation // Proceedings of the 1998 ACM SIGMETRICS International Conference on Measurement and Modeling of Computer Systems. – 1998. – P. 151–160.Barford P., Crovella M. Generating Representative Web Workloads for Network and Server Perfor-mance Evaluation // Proceedings of the 1998 ACM SIGMETRICS International Conference on Meas-urement and Modeling of Computer Systems. – 1998. – P. 151–160.Weinreich H., Obendorf H., Herder E., Mayer H. Not Quite the Average: An Empirical Study of Web Use // ACM Transactions on the Web. – 2008. – Vol. 2, No. 1. – Art. 5. – P. 1–31. – DOI: 10.1145/1326561.1326566Weinreich H., Obendorf H., Herder E., Mayer H. Not Quite the Average: An Empirical Study of Web Use // ACM Transactions on the Web. – 2008. – Vol. 2, No. 1. – Art. 5. – P. 1–31. – DOI: 10.1145/1326561.1326566Zhang Y., Chen W., Wang D., Yang Q. User-click Modeling for Understanding and Predicting Search-Behavior // KDD '11: Proceedings of the 17th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. – San Diego, CA, USA, 2011. – P. 1388–1396. – DOI: 10.1145/2020408.2020613Zhang Y., Chen W., Wang D., Yang Q. User-click Modeling for Understanding and Predicting Search-Behavior // KDD '11: Proceedings of the 17th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. – San Diego, CA, USA, 2011. – P. 1388–1396. – DOI: 10.1145/2020408.2020613Hansen N. The CMA Evolution Strategy: A Tutorial // arXiv preprint arXiv:1604.00772. – 2016.Hansen N. The CMA Evolution Strategy: A Tutorial // arXiv preprint arXiv:1604.00772. – 2016.

The authors declare that there are no conflicts of interest present.