Research Issues and Challenges in AI-Embedded 6G Network Architecture
Anirban Khara, Iqra Javid, Сибарам Хара, Prasad Ramjee
Abstract
G will have extensive features of AI not only in providing AI-based services to users, but also in managing 6G networks making it more intelligent through its embedded AI tools. It will facilitate to implement AI-driven services using massive sensors and IoT devices. Native AI embedded in 6G network architecture will help optimize the spectrum, resources and performance in real-time. AI will make 6 G an autonomous network being self-organizing, self-healing and self-managed. The AI-driven applications are useful in sensing, industry automation, smart city and healthcare. We present the 6G network architecture for different sliced networks like sensing, IoT, distributed computing and satellite integrated 6G networks to identify different network operations that are to be managed by native AI. This also outlines the main technological challenges in 6 G and associated roles of AI for solutions. We present AI deployment scenarios based on various technological issues in 6G. Finally, we emphasize the importance of using appropriate AI tools specific 6 G issues giving examples of mapping between 6 G applications and recommended AI tools. We present the burning issues of AI implementation in 6G and challenges ahead. This work shall be useful for researchers to take up the issues and challenges for development of AI models and testify in the network for implementation.