近年来，随着微机电系统、无线通信和传感器等技术的飞速发展，无线传感器网络的服务能力和服务质量得到了极大的提升，其应用范围也得到了很大的拓展，被广泛应用于战场监视、环境监测、海洋数据采集和入侵检测等领域，目前已成为工业界和学术界的一个研究热点。覆盖问题是无线传感器网络的基本问题，根据覆盖类型的不同，可分为点覆盖、区域覆盖和栅栏覆盖三种。作为无线传感器网络的一个关键问题，栅栏覆盖聚焦于对受监控区域边界的覆盖，目的在于检测试图穿越受监控区域的移动目标。与区域覆盖和点覆盖相比，栅栏覆盖无需覆盖整个区域或全部目标点，而只需覆盖受监控区域的边界，因此，它所需的传感器节点数远远少于前两者，在实际应用中可以更经济地满足应用需求，已被广泛应用于各种入侵检测应用场景。本文面向水下入侵检测应用场景，基于三维的水下无线传感器网络，对水下无线传感器网络中的三维栅栏覆盖问题进行了创新性研究，主要研究内容包括： 1.水下三维栅栏覆盖的高效部署问题。基于水下无线传感器网络的三维栅栏覆盖通常部署在江河、湖泊和海洋等水域，由于其覆盖范围广，涉及的传感器节点多，且目标区域可能有潜在危险导致人类无法接近，因此，通过人工方式将传感器节点逐一部署至目标位置通常不可行。一方面，人工部署的效率低，成本高；另一方面，为保证生命安全，人类不宜靠近危险水域。针对这一问题，本文结合水生环境和水下无线传感器节点的特性，提出了一种部署水下三维栅栏覆盖的高效策略。特别地，针对所提策略中传感器节点的分配问题，设计了一种高效的传感器节点分配算法。仿真结果表明，本文所提算法能高效地将传感器节点分配至目标位置，与经典的匈牙利算法相比，算法计算耗时更少，得到的传感器节点最大移动距离更短，而传感器节点的总移动距离则与之相近。 2.水下三维弱k-栅栏覆盖的判定和构建问题。目前，针对无线传感器网络中的弱k-栅栏覆盖问题，绝大部分研究都假设传感器节点分布在二维平面，相关研究成果被广泛应用于基于二维无线传感器网络的实际应用。然而，由于水下无线传感器网络部署在水下三维环境，其目标部署区域在空间上比二维无线传感器网络多了一个维度，因此，这些基于二维无线传感器网络的研究成果均无法直接应于基于三维水下无线传感器网络的实际应用。针对这一问题，本文结合水下无线传感器网络的特性，通过分析弱k-栅栏覆盖在二维地面无线传感器网络和三维水下无线传感器网络中的异同，提出了一种时间复杂度为O(n²)的水下三维弱k-栅栏覆盖判定算法，该算法可快速判定已部署的水下无线传感器网络能否提供满足实际应用需求的三维弱k-栅栏覆盖。此外，还提出了一种针对水下三维弱k-栅栏构建的传感器节点平行移动方式，并基此设计了一种构建水下三维弱k-栅栏覆盖的高效算法，该算法可以最少的传感器节点构建水下三维弱k-栅栏覆盖，并最小化传感器节点的总移动距离。 3.水下三维强k-栅栏覆盖的构建问题。强k-栅栏覆盖提供k倍于单个栅栏的覆盖强度，确保每一个沿任意路径穿越受监控区域的移动目标都被至少k个传感器节点检测到，可满足应用对栅栏覆盖的更高服务质量要求。鉴于水下三维强k-栅栏覆盖在部署空间上比二维强k-栅栏覆盖多了一个维度，且现有的二维强k-栅栏构建算法均无法直接应用于水下三维强k-栅栏覆盖，本文面向水下三维无线传感器网络的实际应用场景，利用水下传感器节点的可移动性和自身具备的通信能力，设计了一个完全分布式的水下三维强k-栅栏覆盖构建算法，该算法可利用现有的传感器节点构建强度最大的水下三维强k-栅栏覆盖。仿真结果表明，本文提出的算法能保证在有限时间内完成水下三维强k-栅栏构建，所构建的栅栏具备自我修复能力，可在传感器节点受到攻击、突然失效或故障时，进行自我修复并恢复至可达到的最大覆盖强度。此外，该分布式算法在传感器节点的最大移动距离和总移动距离等性能指标方面接近Hungarian和HungarianK这两个集中式算法，而在算法耗时方面则优于后两者。 综上所述，本文基于三维的水下无线传感器网络，面向水下入侵检测应用场景，以构建满足不同应用需求的水下三维栅栏覆盖为目标，提出了水下三维栅栏覆盖的高效部署策略、水下三维弱k-栅栏覆盖判定算法、水下三维弱k-栅栏覆盖构建算法和水下三维强k-栅栏覆盖构建算法，对进一步推进无线传感器网络中三维栅栏覆盖的研究和应用具有一定的理论意义和应用价值。 关键词：无线传感器网络；水下无线传感器网络；栅栏覆盖；三维栅栏覆盖；弱k-栅栏覆盖；强k-栅栏覆盖；构建三维栅栏覆盖

 In recent years, with the rapid development of technologies such as micro-electro-mechanical system, wireless communication and sensor, wireless sensor networks (WSNs) have made great progress in service capabilities and service quality. Now, it has become a research hotspot in industry and academia, and has been widely used in many practical applications such as battlefield surveillance, environmental monitoring, marine data collection and intrusion detection. Coverage problem is an essential problem of WSNs, it can be classified into three categories according to different types of coverage, i.e., point coverage, area coverage and barrier coverage. As a key issue in WSNs, barrier coverage focuses on coverage of the boundaries of the monitored region to detect moving objects that attempt to penetrate the monitored area. Compared with area coverage and point coverage, barrier coverage does not necessarily cover the entire area or al1 target points, but rather only needs to cover the boundary of the monitored region. Therefore, in practical applications, it requires much less sensor nodes than the first two, and can meet the application requirements more economically. Based on three-dimensional (3D) underwater wireless sensor networks (UWSNs), this paper aims at innovative research on 3D underwater barrier coverage in the context of underwater intrusion detection application scenarios. The main contributions are summarized as follows: 1.Efficient deployment of 3D underwater barrier coverage. The UWSNs-based 3D barrier coverage is typically deployed in regions such as rivers, lakes, and oceans. It is generally not feasible to manually deploy sensor nodes one by one to their target positions thanks to the wide range and potential dangerous of the deployment region. On the one hand, manual deployment is costly and inefficient; on the other hand, in order to ensure life safety, human beings should not approach dangerous waters. To address this problem, this paper proposes an efficient deployment scheme for 3D underwater barrier coverage based on the special characteristics of aquatic environment and under-water wireless sensor nodes. In particular, an efficient matching algorithm is devised for the assignment of sensor nodes in the proposed scheme. Simulation results show that the proposed algorithm can efficiently assign sensor nodes to their target positions. Compared with the classical Hungarian algorithm, the proposed algorithm takes much less time, the maximum movement distance of the sensor nodes is shorter, and the total movement distance of all sensor nodes is close to it. 2.Determination and construction of 3D underwater weak k-barrier coverage. At present, most work on weak k-barrier coverage assumes that sensors are distributed on 2D plane, relevant achievements of which are widely used in practical applications based on 2D WSNs. However, compared to traditional 2D WSNs, UWSNs, which are deployed in 3D underwater environment, needs to consider one more dimension. Therefore, the achievements based on 2D assumption cannot directly be applied to practical applications based on 3D UWSNs. This paper combines the characteristics of UWSNs to analyze the similarities and differences between 2D terrestrial weak k-barrier coverage and 3D underwater weak k-barrier coverage, and proposes a decision algorithm of time complexity of O(n²), which can quickly determine whether a deployed UWSN can provide 3D barrier coverage that meets practical application requirements. In addition, a type of parallel movement for constructing 3D underwater weak k-barrier coverage is proposed, based on which an efficient algorithm is devised to construct 3D underwater weak k-barrier coverage. The proposed algorithm can construct 3D underwater weak k-barrier coverage with minimum number of sensor nodes while minimizing the total moving distance of sensor nodes. 3.Construction of 3D underwater strong k-barrier coverage. Strong k-barrier coverage provides k times coverage strength of a single barrier. It can ensure that each intruder that attempts to penetrate the monitored area along any path is detected by at least k sensor nodes, and thus meet the higher requirement of barrier coverage of the practical applications. Given the fact that 3D underwater strong k-barrier coverage has one more dimension than its 2D case, and the existing achievements based on 2D assumption cannot be directly applied to 3D underwater strong k-barrier coverage, this paper devises a fully distributed deployment algorithm for constructing maximum level underwater strong k-barrier coverage with available mobile sensors in 3D underwater environment by using the mobility of underwater sensor nodes and their own communication capabilities. Simulation results show that the proposed algorithm can guarantee the completion of construction of 3D underwater strong k-barrier coverage in a limited time. The constructed 3D underwater strong k-barrier coverage has the self-healing capability to deal with sudden sensor failures, malfunction and being attacked, and restores to the maximum coverage level. In addition, the proposed algorithm obtains acceptable results close to two centralized algorithms named Hungarian and HungarianK in terms of maximum movement distance and total movement distance of the sensor nodes, and is superior to the latter two in terms of running time of algorithm. In summary, this paper aims to investigate 3D barrier coverage problem in UWSNs. More precisely, based on 3D UWSNs and according to different application scenarios of intrusion detection, this paper proposed corresponding scheme and algorithms to address the relevant problems of 3D underwater barrier coverage such as efficiently constructing 3D underwater barrier coverage, determining the exist of 3D underwater weak k-barrier coverage, constructing 3D weak k-barrier coverage and constructing 3D strong k-barrier coverage, which has certain theoretical significance and application value for further promoting the research and application of 3D barrier coverage in WSNs. Key Words: Wireless sensor networks; underwater wireless sensor networks; barrier coverage; three-dimensional barrier coverage; weak k-barrier coverage; strong k-barrier coverage; constructing three-dimensional barrier coverage