Efficient sensor deployment has been one of the most challenging and interesting research areas. The importance and effectiveness of real-world sensing applications, such as underwater and atmospheric sensing, military applications, health systems, and alert systems, which target specific events, raise the need for adaptable design of Wireless Sensor Networks (WSNs). The main challenge in the design of such networks is the optimal sensor deployment, which helps extend the operational network lifetime. Indeed, by maintaining coverage and connectivity with the least number of active nodes and least communication cost, the operable time of the network is guaranteed to be prolonged. The study of two-dimensional (2D) WSNs introduced a significant advancement to the wireless sensor computing technology for different types of smart environments. Nevertheless, 2D WSNs were not sufficient concerning certain applications that require three-dimensional (3D) design. Previous work focused on the design and analysis of various approaches to cover a 3D field of interest, and expanded existing design from 2D to 3D space. Hence, the complexity of such approaches is a major stumbling block. To alleviate this problem, more efficient solutions for the design of WSNs for 3D space deployment have been introduced. By tessellation of the 3D space, which is one of the proposed solutions, researchers studied the partitioning of the space based on Voronoi tessellation by generating identical space-filling cells. Using space fillers cells, which are represented by polyhedra, to model the sensing range of the sensor nodes is assumed to be an optimal solution since these polyhedra can fill a 3D space without leaving gaps or overlaps among them. In the existing literature, the coverage problem in 3D space is concerned with finding the polyhedron that can best approximate the spherical sensing range and eliminates gaps without scarifying the network connectivity. Therefore, the latter is directly related to the sensor node placement strategy. This book chapter studies various proposed solutions for the design of 3D WSNs, with a focus on coverage and connectivity. More specifically, it presents several space filling polyhedra, including the cube, truncated octahedron, hexagonal prism, and rhombic dodecahedron. Also, it compares all these space filling polyhedra to cover a 3D space.
Top1. Introduction
A wireless sensor network is composed of sensors deployed over a field (or geographical region) of interest. It collects and generates data about specific phenomena, and forwards them to a special node, called the base station (or the sink). Wireless sensor networks (WSNs) have multiple fundamental capabilities due to the functionality of the sensor. A sensor node is a small device that consists of four basic components, which are the sensor, microcontroller, energy supplier, and radio transmitter. These four components coordinate together to achieve the service required by the whole network. The sensors’ characteristics are a blessing as they help in the design process of networks for critical missions. However, due to the sensors’ limited resources, such as power (or energy), this may cause the network to be vulnerable network, if it is not designed to consider optimal resource allocation. From now on, we use the terms “sensor” and “sensor node” interchangeably.
Figure 1.
Typical structure of a wireless sensor network
Sensors can be deployed inside the phenomena to be monitored or close to it, depending on the nature of the events to be detected. The frugal technology for designing and implementing WSNs is an attractive source for a wide range of applications, such as military applications, where sensors can be deployed randomly to detect and monitor military targets, and environmental and home applications, which require deterministic deployment, where the locations of sensors need to be predefined. In addition, acoustic applications are being recently introduced as examples of three-dimensional (3D) WSNs. The features of a sensor node ensure wide range of applications. In particular, the sensors have the ability of detecting events using their sensing functionality, and generating data using computational capabilities. Then, the collected data will be forwarded to the sink for processing purposes. To keep the network efficiently valuable, analysis of data collected takes place in order to determine whether data forwarding is necessary or not. Thus, based on the importance and criticality of the event, the sensor decides whether the data should be forwarded or not, to avoid any unnecessary consumption of the sensors’ energy, thus, helping extending the network lifetime. Other approaches, such as clustering (Attarzadeh, 2004), help also increase prolong the lifetime of the individual sensor nodes and that of the whole network. Liu et al. (2010) presented a scheme for balancing energy consumption among sensor nodes with a goal to maximize the network lifetime of 3D WSNs.
The design of 2D WSNs has been extensively studied (Jin, 2012; Zhao, Jiao, Lei, Wu, & Liu, 2009). Since the height of the network is negligible compared to its length and width, the deployment of sensors in 2D WSNs takes place in space and it can be considered efficient to some extent. However, in the sensing of more critical spaces, such as underwater acoustic, atmospheric regions and amicable environments, the 2D design fails to meet the expectation in terms of collecting accurate results. Therefore, recent work has been focused on the investigation of the design of WSNs in 3D space with a goal to suggest new solutions that meet the needs of the above applications. The design of efficient 3D WSNs is mainly related to the analysis and development of algorithms to achieve full coverage of a targeted area, while optimizing the network resources (Akewa, & Thakur, 2012). Designing computational geometry and grid based algorithms has been proved to be very effective in terms of coverage and connectivity (Byun, 2007). Voronoi tessellation is one of the basic concepts that has been used in the design of such algorithms. This book chapter presents an in-depth investigation of the advantages of these algorithms, which help improve existing sensor placement approaches (Domingo, 2009) to locate sensors using Voronoi diagram (Aurenhammer, 1991). A generalization of Voronoi diagram was used to solve the problem of coverage problem in WSNs, where the sensing range of the sensor nodes is an ellipse (Man-Cho So, & Ye, 2005).