The growing popularity of Internet of Things (IoT) with the requirements of highly reliable and low latency has imposed huge challenges to current cellular networks. Using small aerial platforms like unmanned aerial vehicles (UAVs) to assist terrestrial base stations (BSs) is attractive, but it often challenged by the lack of UAV access selection and resource allocation algorithm to balance the network performance and service cost. In this paper, we study the UAV access selection and BS bandwidth allocation problems in a UAV assisted IoT communication network, where a hierarchical game framework is presented. The complicated interactions among UAVs and BSs as well as the cyclic dependency is studied by applying the Stackelberg game theory. Wherein, the access competition among groups of UAVs is formulated as a dynamic evolutionary game and solved by an evolutionary equilibrium. On the other hand, the problem of how much bandwidth should BSs allocate to the UAVs is modeled as a noncooperative game, where the existence and the uniqueness of Nash equilibrium is analyzed. Stochastic geometry tool is used to model the position distribution of network nodes and drive the payoff expressions by taking into account different network parameters. The analytical results for the proposed hierarchical game model and the corresponding solutions are evaluated via simulations, which verify both the validity of our analysis and the effectiveness of the proposed algorithms.

中文翻译：

---

无人机辅助物联网通信网络中联合接入选择和资源分配的博弈论方法

物联网（IoT）的日益普及以及对高度可靠和低延迟的要求，给当前的蜂窝网络带来了巨大的挑战。使用诸如无人机（UAV）之类的小型空中平台来辅助地面基站（BS）颇具吸引力，但它常常因缺乏无人机访问选择和资源分配算法来平衡网络性能和服务成本而面临挑战。在本文中，我们研究了无人机辅助物联网通信网络中的无人机访问选择和基站带宽分配问题，其中提出了分层的游戏框架。应用Stackelberg博弈理论研究了无人机与基站之间复杂的相互作用以及循环依赖关系。其中，无人机群之间的进入竞争被公式化为动态进化博弈，并通过进化均衡来解决。另一方面，将基站应分配给UAV多少带宽的问题建模为非合作博弈，其中分析了纳什均衡的存在和唯一性。随机几何工具用于对网络节点的位置分布进行建模，并通过考虑不同的网络参数来驱动收益表达。通过仿真评估了所提出的分层博弈模型和相应解决方案的分析结果，这验证了我们分析的有效性和所提出算法的有效性。BS应分配给UAV多少带宽的问题被建模为非合作博弈，其中分析了纳什均衡的存在和唯一性。随机几何工具用于对网络节点的位置分布进行建模，并通过考虑不同的网络参数来驱动收益表达。通过仿真评估了所提出的分层博弈模型和相应解决方案的分析结果，这验证了我们分析的有效性和所提出算法的有效性。BS应分配给UAV多少带宽的问题被建模为非合作博弈，其中分析了纳什均衡的存在和唯一性。随机几何工具用于对网络节点的位置分布进行建模，并通过考虑不同的网络参数来驱动收益表达。通过仿真评估了所提出的分层博弈模型和相应解决方案的分析结果，这验证了我们分析的有效性和所提出算法的有效性。