Abstract: Aiming at the problem of energy limitation of Unmanned Aerial Vehicle (UAV) in secure data collection scenario, a secure and efficient data collection optimization scheme based on UAV was proposed to maximize the Safety and Energy Efficiency (SEE) of the system by jointly optimizing the channel relationship, the transmit power of ground sensors and the flight trajectory of UAV. Because the objective function has the characteristic of fractional optimization, it is difficult to solve the problem directly. In order to solve the above problem, an iterative algorithm based on alternating optimization was designed. The auxiliary variable was introduced and the Taylor series expansion method was used to solve the optimization problem of channel relationship and ground sensor node transmit power, and the reinforcement learning method was used to optimize the UAV trajectory. The performance of the proposed scheme was evaluated and compared with the secrecy rate of the scheme with propulsion power limit, without propulsion power limit and fixed transmit power. The experimental results show that the security energy efficiency of the proposed scheme is increased by 9.6%, 37.3% and 27.7% respectively under different flight times, and the system secrecy rate is improved, and the propulsion power is effectively saved.

Key words: Unmanned Aerial Vehicle &#40, UAV&#41, data collection, eavesdropping

摘要： 针对无人机(UAV)在安全数据收集场景中的能量受限问题，通过联合优化信道关系、地面传感器发射功率和UAV飞行轨迹，以最大化系统安全能效(SEE)为目标，提出一种基于UAV的安全高效数据收集优化方案。由于目标函数具有分式优化特性，使问题难以直接求解。为解决上述难题，设计了一种基于交替优化的迭代算法，通过引入辅助变量并利用泰勒级数展开方法求解信道关系和地面传感器节点发射功率的优化问题，同时采用强化学习方法优化UAV轨迹。对所提方案性能进行评估，并与有推进功率限制、无推进功率限制以及固定发射功率方案下的安全保密率进行比较。实验结果表明，在不同的飞行时间下，所提方案的安全能效分别提高了9.6%、37.3%和27.7%，同时提高了系统保密率，并有效节省了推进功率。

关键词: 无人机, 数据收集, 窃听, 安全能效, 迭代优化