The real-time computing model for a network based control system

This paper studies a network based real-time control system, and proposes to model this system as a periodic real-time computing system. With efficient scheduling algorithms and software fault-tolerance deadline mechanism, this model proves that the system can meet its task timing constraints while tolerating system faults. The simulation study shows that in cases with high failure probability, the lower priority tasks stiffer a lot in completion. rate. In cases with low failure probability, this algorithm works well with both high priority and lower priority task. This conclusion suggests that a Internet based control system should manage to keep the failure rate to the minimum to achieve a good system performance.