通信学报2026，Vol.47Issue(3)：195-208,14.DOI:10.11959/j.issn.1000-436x.2026056
多指标意图驱动的无人机计算卸载与轨迹规划自适应优化策略
Multi-indicator intention-driven adaptive optimization strategy for UAV computation offloading and trajectory planning
摘要
Abstract
Unmanned aerial vehicle(UAV)-assisted mobile edge computing(MEC)systems have proven to provide an effective solution to improve the performance of mobile network services and alleviate the shortage of computing power.Targeting complex low-altitude service scenarios,a multi-indicator intent-driven strategy was proposed for UAV compu-tation offloading and trajectory planning.The impact of dynamic changes in service delay,energy consumption and load balancing on system decisions was comprehensively considered,and a joint optimization problem with dynamic multi-indicator intents as the objective was constructed.The original problem,which was characterized by dynamic uncertainty of target parameters,high dimensionality of the solution space and difficulty in direct solution,was decomposed into two subproblems,namely computation offloading and resource allocation,and trajectory planning.A hybrid algorithm inte-grating meta-reinforcement learning and successive convex approximation was designed to realize the alternating optimi-zation of the two subproblems.Simulation experiments were carried out to evaluate the convergence of the proposed al-gorithm in dynamic multi-objective optimization,and the optimization objectives were compared with those of existing schemes.Experimental results show that the proposed algorithm can quickly adapt to the dynamic changes of multiple in-dicators,and significantly reduce system energy consumption and service delay while balancing the UAV load.关键词
计算卸载/负载均衡/轨迹优化/资源分配/元强化学习Key words
computation offloading/load balancing/trajectory optimization/resource allocation/meta-reinforcement learning分类
信息技术与安全科学引用本文复制引用
林鹏,黄新梁,宁兆龙,刘艳,郭磊,张治中..多指标意图驱动的无人机计算卸载与轨迹规划自适应优化策略[J].通信学报,2026,47(3):195-208,14.基金项目
国家自然科学基金资助项目(No.62201271,No.62303232) The National Natural Science Foundation of China(No.62201271,No.62303232) （No.62201271,No.62303232）
