Author(s)

I. Žutautaitė-Šeputienė, J. Augutis & E. Ušpuras

Abstract

The processes of network nodes’ resistance/immunity to hazard spread were analyzed in this study. The case when nodes’ immunity is non-constant (which depends on various factors) is analyzed. The mathematical algorithm (based on the Bayesian approach) for non-constant network nodes’ immunity updating by observed data is presented in this paper. A numerical experiment was performed to illustrate the application of the developed algorithm. Keywords: Bayesian approach, non-constant immunity modelling. 1 Introduction More and more attention is focused on the problems related to quantitative assessment of the behaviour of systems/components/infrastructures in terms of dependability and security and, more generally, quality of service indicators. People, technical equipment, computers (hardware, software), etc or their various combinations form a structure network system (i.e. nodes and network lines/channels) (fig. 1). A hazard/error, its propagation and negative effect, the natural physical ageing of the system, the influence of external factors to the system vitality and so on affect the network system dependability and quality of the service. On the other hand, in a network system some (or all) nodes can have immunity from hazards (for example, in computers we have firewalls, antivirus programs against hacker attacks and computer viruses), according to the hazard propagation immunity of the system node changes (which could increase). When analyzing network systems under the influence of the hazard propagation process, one point of interest is the modelling of node resistance

Keywords

Bayesian approach, non-constant immunity modelling.