Abstract
Routing optimization has long been a problem in the networking field. With the rapid development of user applications, network traffic is continuously increasing in dynamicity, making optimization of the routing problem NP-hard. Traditional routing algorithms cannot ensure both accuracy and efficiency. Deep reinforcement learning (DRL) has recently shown great potential in solving networking problems. However, existing DRL-based routing solutions cannot process the graph-like information in the network topology and do not generalize well when the topology changes. In this paper, we propose AutoGNN, which combines a GNN and DRL for the automatic generation of routing policies. In AutoGNN, the traffic distribution in the network topology is processed by a GNN, while a DRL framework is used to train the parameters of neural networks without human expertise. Our experimental results show that AutoGNN can improve the average end-to-end delay of the network by up to 19.7% as well as present more robustness against topology changes.
Highlights
With the rapid development of new information technologies, such as virtual reality (VR), 4K+ video, online conferences, and cloud services, among others, the information system infrastructure has recently come under a great burden of traffic transmission
All the Deep reinforcement learning (DRL)-based schemes mentioned above used a new method for missing data estimation or a simple multi-layer perceptron (MLP) to carry out the calculation of the network traffic
The DRL Framework of AutoGNN we introduce the mechanism of AutoGNN and the corresponding technologies and algorithms used in AutoGNN
Summary
With the rapid development of new information technologies, such as virtual reality (VR), 4K+ video, online conferences, and cloud services, among others, the information system infrastructure has recently come under a great burden of traffic transmission. Supervised learning uses a labelled data set to train the target algorithm in order to obtain an accurate model for a target problem These methods usually achieve good performance in terms of the accuracy of classification of the input data, but a labelled data set is difficult to acquire in practice in the scenario of routing or traffic engineering. Such methods usually cannot reach good accuracy Both supervised learning and unsupervised learning methods only achieve primitive classification of the network traffic, requiring the use of other algorithms to make a step forward for the generation of routing or traffic engineering policies. To solve the problems related to the application of machine learning technologies in networks, in this paper, we combine the DRL and graph neural network (GNN) frameworks to generate routing policies for the communication networks.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
