Earliest deadline first scheduling technique for different networks in network control system

The idea of this research is to determine the best network providing the reduced network time delay in network controlled system (NCS). NCS tackles robust and nonlinear processes that are time-varying or the requirements for the control change with the operating conditions using earliest deadline first (EDF) scheduling technique. It is very useful for minimizing the effects of parameter variations with respect to the process dynamics. The main advantage of applying EDF is that the parameters of the controller can be changed according to deadline along with the process changes. The main limiting factor is that auxiliary measurement should respond quickly to the process change. By applying the EDF algorithm in NCS schemes, the control of the two-tank water-level system is simulated using PID controllers and the control via Internet is simulated with the constant time delays and time-varying delay conditions. The drawbacks of NCS are network-induced delay, packet losses, bit errors, time-varying random transmission delay, lack of synchronization between devices, bandwidth limitations and integration of the modules. As the EDF scheduling approach is based on the mean time-delay parameters, they shift from one set to another when the mean time delay changes from one set to another. Therefore, time-delay issues are reduced by this algorithm. In this research work, the main challenge network-induced delay is reduced by scheduling technique and suitable network for PID control for EDF scheduling is determined among CAN, Ethernet, Switched Ethernet, TDMA and FDMA. CAN and Ethernet have the minimum of 1ms network-induced time delay when compared with Switched Ethernet, TDMA and FDMA under overload condition. The integral of absolute error is employed to measure the quality of control of each control loop value for EDF scheduling 0.00354 and processor utilization is estimated as U=0.12.