Research on Virtual Network Embedding Model and Algorithm Based on Graph Attention Network and Multi-Layer Perceptron
Hongyan Liu · 2023
Currently, 5G networks are widely used, and 6G network research has attracted much attention, but the development of hardware networks is slow, and network virtualization techniques need to be studied to achieve resource sharing, in which the virtual network embedding problem is the key challenge. In this paper, a virtual network embedding algorithm model with graph attention mechanism and multi-layer perceptron is proposed based on graph attention mechanism and multi-layer perceptron, which is used to determine the weights between nodes by introducing the attention mechanism to measure the relationship between nodes, and through simulation experiments, it can be verified that the models and algorithms proposed in this paper effectively reduce the degree of resource fragmentation, and at the same time improve the acceptance rate of virtual network requests.