Resource-oriented approaches to implementation of traffic control technologies in safety-critical I&C systems

Abstract

Changes in approaches to safety-critical Instrumentation and Control (I&C) systems traffic control results in changes in system power consumption. Thus, timely detection of traffic anomalies allows significant reducing of power consumption due to effective redistribution of the data flows. Well-balanced traffic allows reducing the power consumption of the network compared to the same network, but without such traffic balancing. In this chapter we consider appropriate approaches intended to solve the problem of traffic control in safety-critical I&C systems in order to reduce data flows transmission time. Method of rapid identification of traffic parameters on the basis of the current samples for timely detection of traffic anomalies is proposed. The method allows reducing time of traffic main parameters identification by 15–20%. This makes it possible to determine the behavior of the existing traffic. Moreover, method for short-term prediction of traffic behavior is also proposed. Standard tools are used for flexible and poorly aggregated traffic. An original approach is proposed for the prediction of fractal traffic. Such approach is based on certain properties of fractal processes, including self-similarity and linearity of autocorrelation function. The results of the prediction are used for dynamic control of load distribution in virtual connections, and methods of its effective redistribution allow reducing system power consumption. It is significant that minimal power consumption was achieved when alternative routes were used to reroute the traffic, which was previously flowing through network bottlenecks. Simulation results proved that proposed approach allowed reducing power consumption during appearance anomalies in traffic behavior, compared to standard approaches, by approximately 20%. Moreover, even if traffic behavior is free from anomalies, proposed approach allows increase energy efficiency of the system. Thus, due to well-balanced load caused by data flows into physical links, power consumption of safety-critical I&C systems can be reduced by 7%.