Search

Firstly, the basic concept of network layer hopping technology was introduced and the security threats it faced were given. Then, two type of models and communication methods of network layer hopping technology based on traditional networks and software-defined networks were given. And the network layer hopping technology was classified from three aspects of hopping attributes, the method of hopping realization and the method of hopping trigger, two evaluation models of network layer hopping were given. Finally, the problems that still exist in the network layer hopping technology and the corresponding development direction were summarized.

Container technology promotes the rapid development of cloud computing with its flexible and efficient features, but it also introduces security threats such as co-resident attacks, escape attacks, and common mode attacks. In response to these security threats, a dynamic heterogeneous scheduling method based on Stackelberg game in the container cloud was proposed. First, a heterogeneous mirrored resource pool is constructed to suppress the spread of attacks based on common-mode vulnerabilities on the cloud. Then, the offensive and defense interaction process is modeled as a Stackelberg game model. Finally, the offensive and defensive model is analyzed, and the system scheduling problem is modeled as a mixed integer non-linear programming problem to solve the system's optimal scheduling strategy. Experiments show that the proposed method can improve the defense effect of the cloud platform and reduce the system defense overhead.

Adversarial training suffers from poor effectiveness due to the challenging optimisation of loss with hard labels. To address this issue, adversarial distillation has emerged as a potential solution, encouraging target models to mimic the output of the teachers. However, reliance on pre‐training teachers leads to additional training costs and raises concerns about the reliability of their knowledge. Furthermore, existing methods fail to consider the significant differences in unconfident samples between early and late stages, potentially resulting in robust overfitting. An adversarial defence method named Clean, Performance‐robust, and Performance‐sensitive Historical Information based Adversarial Self‐Distillation (CPr & PsHI‐ASD) is presented. Firstly, an adversarial self‐distillation replacement method based on clean, performance‐robust, and performance‐sensitive historical information is developed to eliminate pre‐training costs and enhance guidance reliability for the target model. Secondly, adversarial self‐distillation algorithms that leverage knowledge distilled from the previous iteration are introduced to facilitate the self‐distillation of adversarial knowledge and mitigate the problem of robust overfitting. Experiments are conducted to evaluate the performance of the proposed method on CIFAR‐10, CIFAR‐100, and Tiny‐ImageNet datasets. The results demonstrate that the CPr&PsHI‐ASD method is more effective than existing adversarial distillation methods in enhancing adversarial robustness and mitigating robust overfitting issues against various adversarial attacks. [ABSTRACT FROM AUTHOR]

Defensive deception is emerging to reveal stealthy attackers by presenting intentionally falsified information. To implement it in the increasing dynamic and complex cloud, major concerns remain about the establishment of precise adversarial model and the adaptive decoy placement strategy. However, existing studies do not fulfil both issues because of (1) the insufficiency on extracting potential threats in virtualisation technique, (2) the inadequate learning on the agility of target environment, and (3) the lack of measurement for placement strategy. In this study, an optimal defensive deception framework is proposed for the container based‐cloud. The System Risk Graph (SRG) is formalised to depict an updatable adversarial model with the automatic orchestration platform. Afterwards, a Deep Reinforcement Learning (DRL) model is trained based on SRG. The well‐trained DRL agent generates optimal placement strategies for the orchestration platform to distribute decoys and deceptive routings. Lastly, the coefficient of deception, C, is defined to evaluate the effectiveness of placement strategy. Simulation results show that the proposed method increases C by 30.22%, and increase the detection ratio on the random walker attacker and persistent attacker by 30.69% and 51.10%, respectively. [ABSTRACT FROM AUTHOR]

To solve the problem that a diversity system is difficult to take defense capability, defense cost and quality of service into account at the same time under a single diversity strategy, firstly, the scheduling object selecting sequences under different security levels were constructed based on the measurement of scheduling heterogeneity, executor security and spatial diversity. Then, according to the coarse-grained evaluation of threat environment, the scheduling time and scheduling object were determined comprehensively. Through the realization of the spatio-temporal diversity Web server system in a cloud environment, the proposed scheduling strategy was tested with attack and defense experiments and compared with the existing scheduling strategies. The results show that the proposed scheduling strategy improves the defense capability significantly and maintains a high quality of service within the acceptable defense cost increasing range. [ABSTRACT FROM AUTHOR]