Reinforcement Learning-Based Service-Oriented Dynamic Multipath Routing in SDN

Figure 9

Operation of a traditional DRL model.