SECURITY MODEL OF MOBILE MULTI-AGENT ROBOTIC SYSTEMS WITH COLLECTIVE MANAGEMENT

The paper deals with creation problem of protection mechanisms for multi-agent robotic systems from attacks by introduced robots-saboteurs. We considered a class of so-called "soft" attacks that involve intercepting of communications, formation and transmission of misinformation to robots group, as well as performing other actions that do not have identified signs of robots-saboteurs invasion. We proposed theoretical security model for multi-agent robotic systems, based on zone security model and model of police stations for distributed computing systems. The basic idea of the proposed subject-object model of access control, is that a logically self-contained entity, the police station, is introduced in information system, in addition to the entities “subject and object”. In accordance with the concept of security monitoring of appeals, it performs the functions of access legitimacy checking and/or integrity of the transactions spatially distributed within a region of subjects and objects. Thus, initially homogeneous multi-agent system is proposed to be designed as heterogeneous, where there are not only agents-executors, but also agents, intended solely for the decision of security problems: identification and authentication, access control, generation, and key distribution and location analysis of agents’ position. For the latter problem solution, the region is divided into several zones with introducing of zonal and interzonal security procedures.  The performance of the model is illustrated by an example of its usage in the protection mechanism creation for classical iterative task of robot forces distribution for several purposes. We show the order of agents’ interaction with the police stations of their zone, as well as implementation of interzonal security policy.