Considering the high delay and packet loss ratio caused by vehicular dynamic topology in pure vehicle-to-vehicle (V2V) communication, vehicle-to-roadside unit (V2R) communication is an important strategy to improve the performance of vehicular ad hoc networks (VANETs). However, it still remains a challenge for vehicles to select the destination uplink node among roadside units (RSUs) deployed in VANETs due to complex and dynamic influence factors such as RSU coverage, traffic density, and vehicle moving distance during uplink. To establish a benchmark for selection of RSUs, we first analyze the data delivery delay quantitatively. Based on a V2R communication classification approach, we estimate the data delivery delay caused by dynamic influence factors. Second, we propose an adaptive V2R communication strategy to provide the most suitable routing choice for minimizing the delay. We also analyze the complexity of the proposed algorithm. Extensive simulations via Network Simulator 2 (NS-2) and NS Scenarios Generator 2 (NSG2) show that proposed algorithm outperforms two basic algorithms, i.e., the shortest distance selection and random selection algorithms and two related methods, i.e., trade-off method and predicted connection method in data delivery delay and packet delivery ratio.

考虑到纯车对车 (V2V) 通信中车辆动态拓扑导致的高延迟和丢包率，车对路单元 (V2R) 通信是提高车辆自组织网络 (VANET) 性能的重要策略）。然而，由于RSU覆盖范围、交通密度和上行期间车辆移动距离等复杂的动态影响因素，车辆在VANET中部署的路侧单元（RSU）中选择目的地上行节点仍然是一个挑战。为了建立选择 RSU 的基准，我们首先定量分析数据传输延迟。基于V2R通信分类方法，我们估计了动态影响因素引起的数据传输延迟。第二，我们提出了一种自适应 V2R 通信策略，为最小化延迟提供最合适的路由选择。我们还分析了所提出算法的复杂性。通过 Network Simulator 2 (NS-2) 和 NS Scenarios Generator 2 (NSG2) 进行的大量模拟表明，所提出的算法优于两种基本算法，即最短距离选择和随机选择算法以及两种相关方法，即权衡方法和数据传输延迟和数据包传输率的预测连接方法。