Deep Edge Node enables AI native toward 6G.

AI native as the key enabler for next-generation (six-generation) communication network has been attracting attention from both academia and industry, as an increasing number of applications will require the intelligence services provided by the edge node. To this end, we present the deep edge node architecture in this paper, which integrates edge computing, network abilities, and AI native. Under this architecture, we also propose the deep reinforcement learning-based approach for the RAN slice problem. Under a simulation environment, the results show that our proposed SAC-based method can reduce latency and improve throughout.