Intelligence slicing: A synthesized framework to integrate artificial intelligence into 5G networks

Abstract

For best-performing networks from 5G and above, it must support a wide range of needs. It is understood that more transmission, resource assistance and communication systems will be required. Achieving these tasks can be challenging as network infrastructure becomes more complex and massive. A good solution is to incorporate more robust AI technology that has been tested to provide answers ranging from channel prediction to autonomous network management, as well as network security. Today, however, the latest technology to integrate AI into wireless networks is limited to using a unique AI algorithm to solve a specific problem. A comprehensive framework that can fully utilize the power of AI in solving various network problems remains an open problem. Therefore, this paper introduces the idea of the spy pieces on which the AI unit is installed and delivers on one condition. Intelligence units are used to flexibly control the intelligence of AI algorithms with two comprehension strategies to perform different intellectual tasks: 1) Neural network-based channel predictions and 2) Industrial network-based security acquisition, to illustrate this framework.