부산대학교 도서관

In the context of future wireless networks, various emerging applications such as AR/VR/XR, e-Health, live video streaming, and automated vehicles are expected to have the diverse and strict quality of service (QoS) requirements. Meeting these requirements with high reliability, low latency, high security, high throughput, and high speed, will demand mobile service providers to offer programmable solutions for delivering services in multiple ways. Network slicing that can be achieved through the combination of Network-Functions Virtualization (NFV) and Software-Defined Networking (SDN) is one such solution for customized service instances. This paper proposes an artificial intelligence (AI)-based deep learning method using fully connected neural networks to select an appropriate network slice for users from a public dataset. The considered communication slices include enhanced Mobile Broadband (eMBB), large-scale machine-type communication (mMTC), and Ultra-Reliable Low Latency Communications (URLLC). The performance evaluation of the proposed slice recognition algorithm demonstrates its high accuracy for recognizing the best slice for a given service.