image: Figure 1 The DHR architecture based on historical credibility and dissimilarity clustering view more
Credit: ©Science China Press
The DHR architecture breaks the pattern of staticity, similarity, and determinism of cyberspace security information systems through multiple heterogeneities in different spatial-temporal dimensions. The reliability and heterogeneity of executors are the basis and prerequisite for achieving security and diversity in mimic defense, which largely determines the upper security limit of DHR architecture. In this work, we propose a DHR executor selection algorithm based on historical credibility and dissimilarity clustering (HCDC). The executors are classified according to historical credibility and dissimilarity, so that the scheduling module can select executors with high credibility and dissimilarity as service executors as much as possible to improve the system's defense against known or unknown attacks. The simulation results demonstrate that, in comparison to existing methods, the algorithm reduces the attack success rate and average failure rate while increasing system reliability.
The contributions of this work can be summarized as follows.
(1) The traditional DHR architecture is improved by introducing the related definitions of executor historical credibility and dissimilarity, which characterize executor reliability and degree of heterogeneity.
(2) The executors are divided into multiple heterogeneous executor pools with large heterogeneity by using the k-means algorithm. The executors with the largest historical credibility are selected from each pool as the candidate pool, with the historical credibility being dynamically updated by negative feedback control based on the results of the multi-mode adjudicator. The executors are randomly selected from the candidate pool as the set of service executors.
(3) The effectiveness of this algorithm is evaluated by theoretical analysis and simulation experiments, and the attack success rate and average failure rate of this algorithm are compared with existing algorithms.
See the article:
Shao SS, Ji Y M, Zhang W L, et al. A DHR executor selection algorithm based on historical credibility and dissimilarity clustering. Sci China Inf Sci, 2023.
http://engine.scichina.com/doi/10.1007/s11432-022-3635-2
Journal
Science China Information Sciences