Figure 3. The multi-agent system-Optimization of Distributed Systems Using Multi-Agent Systems with Virtual Time

<p>The authority is given to a single agent<br> that is symbolically situated on the top level; this agent would be considered the root of an<br> arborescent structure; but the stages of the process are provided by groups of agents, respectively by<br> the relations of communication between them.<br> Thus, we would deal with a ZERO AGENT and several groups of agents. Such a group is<br> made up of several agents, each and one of these agents accomplishing a certain role in the process<br> of synchronization (figure 3).</p>