As a future leading technology, sixth generation (6G) networks should be capable of dynamic allocation of resources, process signals, and change in traffic flow. AI/ML can bring about such solutions to manage such tasks. AI-empowered techniques can be used in optimizing the performance of the network efficiently together with mobile edge computing. AI/ML techniques provide the possibility of producing automatic-learning models for optimized network for 6G wireless networks, which grants operators/providers the access for optimizing parameters of network and automatic network adjustment. In this chapter, the authors explore the various applications of AI/ML in sixth-generation and next generation networks and provide detailed explanation on how AI/ML may be implemented in 6G network effectively. Moreover, they provide possible issues while implementing AI-empowered 6G networks.
TopIntroduction
Wireless communication systems are the factors which play a vital role in modern society for health, commercial, business as well as entertainment purposes. These technologies have been evolving and as of now, industrial consultants predict the development of Sixth-generation (6G) network. 6G, the superseder of fifth-generation (5G) network, is expected to use higher frequencies than 5G and is anticipated to use lower latency of below 1 ms of end-to-end lag, elevated mobility of up to 1000 km, large frequency bands of around 1THz – 3THz and it supports data rates of 1 terabyte (TB) per second. It enables users to get better QoE and QoS, along with enormous improvement in network performance. Moreover, the combination of selection of frequency and sub-millimeter bring forth relative electromagnetic absorption rates which can lead to improvement of wireless sensing technology to a higher level.
One of the primary elements of 6G network will be the utilization of AI and ML techniques. ML is one of the types of AI which learns and trains itself using historical data through algorithms and makes a decision as well as predicts the new input without being explicitly programmed. ML algorithms can be splitted into three types: 1) Supervised Learning, 2) Unsupervised Learning and 3) Reinforcement Learning. This chapter discusses the applications of AI/ML in 6G network along with its implementation of its architecture.
The 6G and next generation networks are expected to be the required phases of society that may need specific essential qualities (Curtis Watson, 2021) such as
- •
Flexibility – The recent need of the next generation network has epidemically increased with unmanned autonomous vehicles (UAVs) and internet of things (IoT). So, it should be flexible to be able to merge with the developing technologies.
- •
Sustainability – Consumption of energy, lifetime of its battery, and its impacts on the environment should be taken into consideration for the prolongation of uninterruptible communication network.
- •
Intellect – AI-enabled smart network, mobile edge technology and network sensing will boost the network and takes it to new value paradigms.
- •
Trust – The network should meet the privacy and security demands.
The features of the 6G network are given in Figure 1. The requirements and the use of AI/ML enabled 6G network is introduced in the introduction section. Background Section provides the literature survey of the related work. AI/ML enabled 6G Section provides the details on the AI/ML empowered 6G network. Applications of AI enabled 6G Section describes the various applications of 6G combined with AI/Ml technologies. While issues that may occur in implementing 6G network are presented in issues section, some potential solutions are discussed in the solutions and recommendations section. Finally, the chapter is arrived at the conclusion section.
TopBackground
In (K. B. Letaief, 2019), Khaled B Letaief, et. al. discussed about potential technologies for 6G network which can enable various mobile AI applications. Moreover, several AI based methodologies for optimization and network design were presented along with the key trends which will be factors for the evolution of 6G.
In (Razvan-Andrei stoica, 2019), Razvan-Andrei Stoica, et. al. proposed an AI based paradigm shift for wireless communication which can be combined with technologies like (NOMA) and full duplex radio. Finally, they provided a method called randomized incoherent tight frames which can lead to optimized maximum likelihood detection.
In (Karan Sheth, Keyur Patel, 2020), Karan Sheth, et. al. presented various future applications which can be integrated AI enabled 6G like drone communication, preservation of security and privacy, object localization etc. Finally, they discussed on the use cases which shows how different AI techniques can be adopted in the intelligent transport system and provided solutions for issues like low communication overhead cost.