School of Electrical and Computer Engineering, National Technical University of Athens, 15780 Athens, Greece.
Department of Electrical and Computer Engineering, University of New Mexico, Albuquerque, NM 87131, USA.
Sensors (Basel). 2019 Sep 6;19(18):3853. doi: 10.3390/s19183853.
Modern Public Safety Networks (PSNs) are assisted by Unmanned Aerial Vehicles (UAVs) to provide a resilient communication paradigm during catastrophic events. In this context, we propose a distributed user-centric risk-aware resource management framework in UAV-assisted PSNs supported by both a static UAV and a mobile UAV. The mobile UAV is entitled to a larger portion of the available spectrum due to its capability and flexibility to re-position itself, and therefore establish better communication channel conditions to the users, compared to the static UAV. However, the potential over-exploitation of the mobile UAV-based communication by the users may lead to the mobile UAV's failure to serve the users due to the increased levels of interference, consequently introducing risk in the user decisions. To capture this uncertainty, we follow the principles of Prospect Theory and design a user's prospect-theoretic utility function that reflects user's risk-aware behavior regarding its transmission power investment to the static and/or mobile UAV-based communication option. A non-cooperative game among the users is formulated, where each user determines its power investment strategy to the two available communication choices in order to maximize its expected prospect-theoretic utility. The existence and uniqueness of a Pure Nash Equilibrium (PNE) is proven and the convergence of the users' strategies to it is shown. An iterative distributed and low-complexity algorithm is introduced to determine the PNE. The performance of the proposed user-centric risk-aware resource management framework in terms of users' achievable data rate and spectrum utilization, is achieved via modeling and simulation. Furthermore, its superiority and benefits are demonstrated, by comparing its performance against other existing approaches with regards to UAV selection and spectrum utilization.
现代公共安全网络 (PSN) 借助无人机 (UAV) 在灾难事件中提供有弹性的通信范例。在这种情况下，我们提出了一种分布式以用户为中心的风险感知资源管理框架，该框架由静态无人机和移动无人机提供支持。移动无人机由于其重新定位的能力和灵活性，可以获得更大比例的可用频谱，因此与静态无人机相比，可以为用户建立更好的通信信道条件。然而，由于干扰水平的增加，用户对基于移动无人机的通信的潜在过度利用可能导致移动无人机无法为用户提供服务，从而导致用户决策中的风险。为了捕捉这种不确定性，我们遵循前景理论的原则，并设计了用户的前景理论效用函数，该函数反映了用户在向基于静态和/或移动无人机的通信选项进行传输功率投资时的风险感知行为。我们制定了用户之间的非合作博弈，其中每个用户确定其向两种可用通信选择的功率投资策略，以最大化其预期的前景理论效用。证明了纯纳什均衡 (PNE) 的存在性和唯一性，并表明了用户策略向其收敛。引入了一种迭代分布式和低复杂度算法来确定 PNE。通过建模和仿真，实现了所提出的以用户为中心的风险感知资源管理框架在用户可实现数据速率和频谱利用率方面的性能。此外，通过将其性能与其他现有方法（关于无人机选择和频谱利用率）进行比较，证明了其优越性和优势。
