AI-Powered Autonomous Network Slicing for 6G Communication Networks in Smart Cities

The demand to make the migration to communication networks based on Sixth Generation (6G) is motivated by the increasing need to be connected on a hyper-scale, extremely reliable, and ultra-low-latency due to the applications of Smart City, which demand a variety of services with a critical level of reliability and importance to the Internet of Things (IoT). The Fifth Generation (5G) network slicing functionalities used today are based on reactive, static, and manual orchestration, which cannot be used in the dynamically complex nature of 6G. In this paper, an AI-based autonomous network slicing framework is proposed, built on an AI-native architecture. Based on the Deep Deterministic Policy Gradient (DDPG) framework, the entire slice life cycle of Preparation, Planning, and Operation phases is dynamically resolved to a multi-objective optimization problem to maximize network utility, considering the strict compliance with heterogeneous Quality of Service (quality of service) constraints. In this autonomous design, resources are proactively predicted and independently adapted to guarantee superior performance and enhanced resilience. The assessments are calculated to show that the efficiency of the framework is high, as reflected in the low Round-Trip Time (RTT), reduction in packet loss, high availability, and high throughput, and at the same time encourage resource efficiency and the high level of security on 256-bit encryption, which is essential in the implementation of the Smart City.

Перевод пока недоступен