Information models of controlled processes in the computer network

Abstract

Nowadays there are a lot of powerful network management tools which allow in comparatively easy way to implement state monitoring of any network element. As a price these tools consume such resources as CPU, memory and bandwidth. Unfortunately in some cases the cost of monitoring can be inadmissible high that makes administrators looking for more economical techniques of network elements controlling. One of such techniques consists in approximation of controlled processes. Mainly this approach helps to reduce the bandwidth usage via decreasing number of measurements. In order to improve efficiency of the approximation two information models have been suggested in the work. Additionally an evaluating method of the information models has been created.