Commercially well-known drones, unmanned aerial vehicles (UAVs), are increasingly popular with the public and have been widely deployed in diverse applications. However, a drone equipped with tracking, monitoring, or sensing device(s) can illegally collect privacy- and security-sensitive information and intrude restricted areas. Thus, recent literature focuses on the protection of users and restricted areas from an unwanted privacy attack and intrusion caused by the drone. Unlike prior research, however, we fundamentally shift the privacy paradigm from protecting users and restricted areas from a malicious drone into protecting and hiding the sensitive information of a drone from an adversary. In light of these, we propose three privacy-preserving target tracking strategies based on the shortest path, random locations, and dummy locations. The basic idea is to obfuscate the current location of the drone and randomize the trajectory to prevent the adversary from locating and tracking the drone. We also analyze drone privacy in terms of location and trajectory and measure them through entropy-based anonymity, the size of convex-hull, and the number of paths. We conduct extensive simulation experiments using the OMNeT++ for performance evaluation and comparison with stationary and moving target tracking scenarios under three mobility models. The simulation results indicate that the proposed strategies can be a viable approach to track the target while reserving a certain level of location and trajectory privacy.

商业上知名的无人机，无人驾驶飞行器（UAV）越来越受公众欢迎，并已广泛应用于各种应用。但是，配备跟踪、监控或传感设备的无人机可能会非法收集隐私和安全敏感信息并侵入限制区域。因此，最近的文献侧重于保护用户和受限区域免受无人机引起的不必要的隐私攻击和入侵。然而，与之前的研究不同，我们从根本上将隐私范式从保护用户和受限区域免受恶意无人机攻击转变为保护和隐藏无人机的敏感信息免受攻击。鉴于这些，我们提出了三种基于最短路径、随机位置和虚拟位置的隐私保护目标跟踪策略。其基本思想是对无人机的当前位置进行模糊处理，并将轨迹随机化，以防止对手定位和跟踪无人机。我们还根据位置和轨迹分析无人机隐私，并通过基于熵的匿名性、凸包大小和路径数来衡量它们。我们使用 OMNeT++ 进行了广泛的仿真实验，以进行性能评估，并与三种移动模型下的静止和移动目标跟踪场景进行比较。仿真结果表明，所提出的策略可以成为一种可行的方法来跟踪目标，同时保留一定程度的位置和轨迹隐私。我们还根据位置和轨迹分析无人机隐私，并通过基于熵的匿名性、凸包大小和路径数来衡量它们。我们使用 OMNeT++ 进行了广泛的仿真实验，以进行性能评估，并与三种移动模型下的静止和移动目标跟踪场景进行比较。仿真结果表明，所提出的策略可以成为一种可行的方法来跟踪目标，同时保留一定程度的位置和轨迹隐私。我们还根据位置和轨迹分析无人机隐私，并通过基于熵的匿名性、凸包大小和路径数来衡量它们。我们使用 OMNeT++ 进行了广泛的仿真实验，以进行性能评估，并与三种移动模型下的静止和移动目标跟踪场景进行比较。仿真结果表明，所提出的策略可以成为一种可行的方法来跟踪目标，同时保留一定程度的位置和轨迹隐私。