AI-Native Orchestration in the 6G Continuum: Evolving Operator Platforms with Agentic AI
本文提出了一种基于自主智能体的AI原生编排层,用于6G网络中的无缝云-边缘-IoT连续体,以解决跨域服务自动化和管理问题。
本文提出了一种基于自主智能体的AI原生编排层,用于6G网络中的无缝云-边缘-IoT连续体,以解决跨域服务自动化和管理问题。
该研究提出一种结合外部校准相机、对象检测、持续跟踪及本体驱动语义更新的混合管道,以构建动态语义世界模型,解决机器人在复杂环境中的交互问题。
研究解决了工厂助手模型在硬件限制下的部署问题,通过结构压缩、检索增强适应及子网络选择方法优化模型大小、速度和质量。
Deploying AI at the edge demands co-optimization of model accuracy, efficiency, and hardware deployability, yet existing neural architecture search (NAS) methods often neglect systematic consideration of quantization effects and hardware mapping. This work proposes a three-stage joint optimization framework comprising a hardware-agnostic Pareto-ranking proxy front-end, a quantization perturbation-aware Pareto filtering and feedback bridge, and an evolutionary hardware-space exploration back-end tailored for CGRA4ML. For the first time, it systematically characterizes the impact of INT4 post-training quantization across the full Pareto space of NAS-Bench-201 architectures, revealing that FP32 zero-shot proxies outperform dedicated INT4 proxies in Pareto coverage, and enables efficient automatic mapping of quantized models onto reconfigurable accelerators.
This work proposes an efficient real-time nonlinear model predictive control (NMPC) approach for the remote underactuated double pendulum system under unknown parameters and limited interaction time. By integrating a structure-exploiting alternating direction method of multipliers (ADMM) with an interior-point-accelerated sequential quadratic programming (SQP) solver, the method achieves global swing-up and stabilization without requiring prior model knowledge. The proposed framework is directly deployed on the CloudPendulum remote hardware platform, satisfying stringent real-time constraints while demonstrating strong robustness and disturbance rejection capabilities. Experimental results confirm its effectiveness in reliably achieving both swing-up and stable regulation of the double pendulum system.
本文提出了一种基于自主智能体的AI原生编排层,用于6G网络中的无缝云-边缘-IoT连续体,以解决跨域服务自动化和管理问题。
该研究提出一种结合外部校准相机、对象检测、持续跟踪及本体驱动语义更新的混合管道,以构建动态语义世界模型,解决机器人在复杂环境中的交互问题。
研究解决了工厂助手模型在硬件限制下的部署问题,通过结构压缩、检索增强适应及子网络选择方法优化模型大小、速度和质量。
Deploying AI at the edge demands co-optimization of model accuracy, efficiency, and hardware deployability, yet existing neural architecture search (NAS) methods often neglect systematic consideration of quantization effects and hardware mapping. This work proposes a three-stage joint optimization framework comprising a hardware-agnostic Pareto-ranking proxy front-end, a quantization perturbation-aware Pareto filtering and feedback bridge, and an evolutionary hardware-space exploration back-end tailored for CGRA4ML. For the first time, it systematically characterizes the impact of INT4 post-training quantization across the full Pareto space of NAS-Bench-201 architectures, revealing that FP32 zero-shot proxies outperform dedicated INT4 proxies in Pareto coverage, and enables efficient automatic mapping of quantized models onto reconfigurable accelerators.
This work proposes an efficient real-time nonlinear model predictive control (NMPC) approach for the remote underactuated double pendulum system under unknown parameters and limited interaction time. By integrating a structure-exploiting alternating direction method of multipliers (ADMM) with an interior-point-accelerated sequential quadratic programming (SQP) solver, the method achieves global swing-up and stabilization without requiring prior model knowledge. The proposed framework is directly deployed on the CloudPendulum remote hardware platform, satisfying stringent real-time constraints while demonstrating strong robustness and disturbance rejection capabilities. Experimental results confirm its effectiveness in reliably achieving both swing-up and stable regulation of the double pendulum system.