A Study on Reinforced Support Vector Machine Classifier for Periodic and Aperiodic Scheduling in Networked Control Systems

Periodic and aperiodic data communication binds with the time of data transmission in any networked control system. The scheduling of periodic and aperiodic is an essential to maintain the process control and maintenance of the plant. The classification implemented through Support Vector Machine (SVM) in the literature may cause ambiguity in differentiating the packets of periodic and aperiodic communication involved in an application configured for networked control system. The present paper studies on the classification of data packets using SVM and the consequences of incorporating reinforcement learning on SVM. This reinforced SVM reduces the ambiguity in differentiating the packet of periodic and aperiodic communication so that scheduling of the packet will be more effective in the networked control system.