Abstract Artificial intelligence (AI) methods for traffic video analysis have been widely identified as potential solutions for solving hard problems in intelligent transport systems (ITS). To exploit the advantages of AI, dense cameras to monitor the traffic are required to be deployed along the road and at the intersections. The captured videos of these cameras should be back-hauled to the control center, acting as the inputs of the AI methods. To bear such large data traffic load and to cover long transmission range, directional communication technology can be employed, which concentrate the energy of the wireless signal in a specified direction to provide high data rate and long transmission range (up to hundreds of kilometers). In this paper, the communication time extension problem (CTEP) is identified when directional transmission is applied to the dense urban traffic surveillance system, where the wireless signal propagation time approximates the data transmission time. A link distance division-based time division multiple access (LDD-TDMA) protocol is proposed to address the identified CTEP. Firstly, the directional wireless communication links are classified into categories according to the link distance, where nodes located in the same communication ring belong to the same category. Then a link distance aware (LDA)-based slot allocation algorithm is proposed to assign the time slots to the links. The optimal communication rings’ radius is derived in closed form formula, and the minimum average links’ distance is derived. Simulation results show that LDD-TDMA outperforms TDMA by 13.37% when the ring number is 4.

中文翻译：

---

面向密集城市交通监控系统的高效多路访问控制协议

摘要 用于交通视频分析的人工智能 (AI) 方法已被广泛认为是解决智能交通系统 (ITS) 难题的潜在解决方案。为了发挥人工智能的优势，需要在道路沿线和十字路口部署用于监控交通的密集摄像头。这些摄像机捕捉到的视频应该回传到控制中心，作为 AI 方法的输入。为了承载如此大的数据流量负载和覆盖远距离传输，可以采用定向通信技术，将无线信号的能量集中在指定方向，以提供高数据速率和远距离传输（可达数百公里）。在本文中，当定向传输应用于密集的城市交通监控系统时，通信时间扩展问题（CTEP）被识别，其中无线信号传播时间近似于数据传输时间。提出了一种基于链路距离划分的时分多址 (LDD-TDMA) 协议来解决已识别的 CTEP。首先，将定向无线通信链路根据链路距离进行分类，位于同一通信环内的节点属于同一类别。然后提出了一种基于链路距离感知（LDA）的时隙分配算法来为链路分配时隙。以闭式公式推导出最优通信环半径，推导出最小平均链路距离。