A Survey on Deep Learning for the Routing Layer of Computer Network

With recent achievements in deep learning over the past year, many computer and network applications actively used deep learning architectures including convolution neural network and long short-term memory to improve the performance of their approach through them. The computer network used a complex and dynamic system. For example, routing is the main networking tasks in the fields of the communication network and it is widely used to optimize the optimal routing from the original host to the destination host. However, most of the traditional routing protocol is based on the experience of experts. In this paper, we present an overview of deep learning methods for the routing layer in the computer network. Furthermore, this paper discusses reinforcement learning methods about network routing. Finally, we outline a summary of the current state-of-the-art approaches along with some future research directions.