内生智能驱动的6G边缘网络双闭环自愈架构设计

Endogenous Intelligence-Driven Dual-Loop Self-Healing Architecture for 6G Edge Networks

智能技术不断涌现,驱动着通信产业持续升级,向自智的更高阶段演进。内生智能是实现自智网络的关键因素。在未来网络管控体系中,如何编排管理智能能力,实现网络故障管理全流程智能化与自动化成为一个重要问题。为此,设计了一种贯穿故障全生命周期的内生智能驱动的6G边缘网络双闭环自愈架构,提出自愈目标和方案的概念,建立了QoS保障的自愈目标损失最小化模型,以及对应的网络决策机制,从而支持从无线接入网到核心网的跨域管理能力。在可编程边缘网络环境中进行了验证实验,结果显示,自愈闭环控制流程正常运行,所提算法在毫秒级时延内解决了链路拥塞故障,保证了自愈目标的实现并满足QoS指标。

The rapid growth of emergence of intelligent technologies are driving the communications industry to continuously upgrade and evolve to a higher stage of autonomy.Endogenous intelligence is a key factor in realizing Autonomous Networks.In the future autonomous network management system,how to orchestrate the management of intelligent capabilities to realize the whole process of network fault management intelligence and automation?To this end,this paper designs an endogenous intelligence-driven 6G edge network dual-loop self-healing architecture throughout the whole fault lifecycle,proposes the concept of self-healing goal and solution,and establishes a loss minimization model of self-healing goal with QoS guarantee,and the corresponding network decision-making mechanism,so as to support the cross-domain management capability from the radio access network to the core network.We conducted validation experiments in a programmable edge network environment,and the results show that the self-healing closed-loop control process operates normally,and the proposed algorithm solves the link congestion faults within milliseconds delay,ensures the realization of the self-healing objective and satisfies the QoS metrics.