Abstract: The present disclosure discloses an intelligent mapping method for a cloud tenant virtual network based on a reinforcement learning model, where a mapping algorithm is used to combine a resource abstraction model, Blocking Island, with a deep reinforcement learning algorithm, Actor-Critic, reasonably abstract underlying network resources by means of the Blocking Island model, and efficiently represent resource connectivity information of the entire network with an amount of available resources between nodes as a lower bound. The method specifically includes: (1): completing modeling of virtual network embedding; (2): modeling computing resources and bandwidth resources in a physical network; (3): constructing a neural network; and the like. Compared with the prior art, the present disclosure has better performance in average mapping cost, benefit-cost ratio, total benefit value and mapping success rate, further improves mapping accuracy, reduces the average mapping cost, and has a wide application prospect.

Abstract: A method for automatically regulating an explicit congestion notification (ECN) of a data center network based on multi-agent reinforcement learning is provided. The method specifically includes steps 1 to 3. In step 1, an ECN threshold regulation of a data center network is modelled as a multi-agent reinforcement learning problem. In step 2, an independent proximal policy optimization (IPPO) algorithm in multi-agent reinforcement learning is used for training according to features of the data center network. In step 3, offline pre-training is combined with online incremental learning such that a model deployed on each switch is capable of rapidly adapting to a dynamic data center network environment.