Smith predictor controller design for TCP/AQM

Congestion is a major issue in any communication network. To alleviate this clog in the AQM (Active Queue Management) topology for TCP/IP networks is recommended. This structure’s goal is to reduce congestion by minimizing the size of the typical queue at the routers. Recently, several feedback strategies were developed to raise AQM. In this study, a fluid-flow model of TCP networks was used to create a Smith predictor controller designed for TCP/AQM. The advantage of using this controller is its capacity to predict a dynamic system’s state and thus reduce the effect of input delay inherent in TCP/IP flows. Smith predictor configuration is designed to attain reasonable process control with a long-dead time. The rising time, peak time, and settling time of the system are used to analyze its performance in the time domain. This designed controller contains two controllers proportional integral $${G}_{c1}$$ G c 1 and proportional derivative $${G}_{c2}$$ G c 2 for different purposes, specifically, the rejection of load disturbances and the time delay while simultaneously stabilizing the unstable process. The designing procedure for controllers $${G}_{c1}$$ G c 1 and $${G}_{c2}$$ G c 2 are self-determining. To provide robustness for process boundary uncertainties, a first-order filter is used. Simulation results demonstrate that the proposed technique’s performance is satisfactory and has improved robustness. It can be observed that the proposed framework can function in noisy environments.