Han Guangjie, Li Shanshan, Zhu Chunsheng, Jiang Jinfang, Zhang Wenbo
Department of Information and Communication Systems, Hohai University, 200 North Jinling Road, Changzhou 213022, China.
Department of Electrical and Computer Engineering, The University of British Columbia, Vancouver, BC V6T 1Z4, Canada.
Sensors (Basel). 2017 Feb 8;17(2):316. doi: 10.3390/s17020316.
Marine environmental monitoring provides crucial information and support for the exploitation, utilization, and protection of marine resources. With the rapid development of information technology, the development of three-dimensional underwater acoustic sensor networks (3D UASNs) provides a novel strategy to acquire marine environment information conveniently, efficiently and accurately. However, the specific propagation effects of acoustic communication channel lead to decreased successful information delivery probability with increased distance. Therefore, we investigate two probabilistic neighborhood-based data collection algorithms for 3D UASNs which are based on a probabilistic acoustic communication model instead of the traditional deterministic acoustic communication model. An autonomous underwater vehicle (AUV) is employed to traverse along the designed path to collect data from neighborhoods. For 3D UASNs without prior deployment knowledge, partitioning the network into grids can allow the AUV to visit the central location of each grid for data collection. For 3D UASNs in which the deployment knowledge is known in advance, the AUV only needs to visit several selected locations by constructing a minimum probabilistic neighborhood covering set to reduce data latency. Otherwise, by increasing the transmission rounds, our proposed algorithms can provide a tradeoff between data collection latency and information gain. These algorithms are compared with basic Nearest-neighbor Heuristic algorithm via simulations. Simulation analyses show that our proposed algorithms can efficiently reduce the average data collection completion time, corresponding to a decrease of data latency.
海洋环境监测为海洋资源的开发、利用和保护提供了至关重要的信息和支持。随着信息技术的快速发展，三维水下声学传感器网络（3D UASNs）的发展为方便、高效、准确地获取海洋环境信息提供了一种新策略。然而，声学通信信道的特定传播效应导致随着距离增加信息成功传递概率降低。因此，我们研究了两种基于概率邻域的数据收集算法，用于3D UASNs，它们基于概率声学通信模型而非传统的确定性声学通信模型。使用自主水下航行器（AUV）沿设计路径遍历，从邻域收集数据。对于没有先验部署知识的3D UASNs，将网络划分为网格可使AUV访问每个网格的中心位置进行数据收集。对于预先知道部署知识的3D UASNs，AUV只需通过构建最小概率邻域覆盖集访问几个选定位置，以减少数据延迟。否则，通过增加传输轮数，我们提出的算法可在数据收集延迟和信息增益之间进行权衡。通过仿真将这些算法与基本的最近邻启发式算法进行比较。仿真分析表明，我们提出的算法可有效减少平均数据收集完成时间，相应地降低数据延迟。
