隨著行動多媒體服務迅速的發展，尤其是超高畫質影像串流和虛擬實境的出現，前瞻多媒體應用，如元宇宙 (Metaverse) 等，逐漸成為行動服務的主流方向，議題的討論也越來越多。隨之而來的挑戰是，串流影像的同時嚴格維持超高畫質影像品質。也就是說，無線網路必須提共更低的延遲並擁有更高傳輸率的服務，這也是下一代行動技術持續討論的議題。為此，第三代夥伴計畫 (3rd Federation Partnership Project，3GPP) 的 5G 標準中定義了許多的功能和選項給更強大且更有彈性的無線網路環境，新的規格可以讓系統在流量、延遲以及可靠性三方面有更好的表現，但是同時也讓無線資源管理的複雜度直線上升。由於複雜度的上升，現有的傳統資源管理方法效果有限，所以我們希望透過增強式學習方法來解決 5G/6G 標準下無線網路資源管理的問題，並實現示範應用場域。。;With the rapid development of mobile multimedia services, especially the emergence of ultra-high-definition video streaming and virtual reality, forward-looking multimedia applications, such as Metaverse, have gradually become the mainstream direction of mobile services, and the discussion of the topic has become more and more more and more. The challenge that comes with it is to strictly maintain ultra-high-definition image quality while streaming. That is to say, wireless networks must provide services with lower latency and higher transmission rates, which are also ongoing discussions on next-generation mobile technologies. To this end, the 5G standard of the 3rd Generation Partnership Project (3GPP) defines many functions and options for a more powerful and resilient wireless network environment. The three aspects of delay and reliability have better performance, but at the same time, the complexity of wireless resource management has skyrocketed. Due to the increase in complexity, the existing traditional resource management methods have limited effect, so we hope to solve the problem of wireless network resource management under the 5G/6G standard through the reinforcement learning method, and realize the demonstration application field.
