Sensor Localization in Three-Dimensional Space: A Survey

Introduction

Wireless sensor networks are becoming increasingly pervasive in various fields of applications thanks to key advances in wireless communication and electronics. The design and development of such type of network is useful to numerous real-world applications, including civilian (health and environment monitoring), natural (disaster detection), military (battlefield surveillance), and agricultural (precision agriculture) applications, to name a few. They are deployed to monitor a specific phenomenon in a given environment and gather the corresponding sensed data. A wireless sensor network consists of low-cost, battery-powered tiny devices, called sensor nodes (or simply sensors), which possess storage, processing, and communication capabilities. Furthermore, these sensors are capable of sensing the environment and collecting data with respect to several parameters, such as temperature, light, sound, vibration, etc. However, because of their tiny physical size, the sensors suffer from their limited capabilities, and particularly, their battery power (or energy). Indeed, usually, these sensors are equipped with AA/AAA batteries with short lifetime. Thus, the main challenge in the deployment of this type of network is scarce energy of the sensors, which is a crucial component for the correct operation of the entire network. These sensors can transmit their sensed data to a central gathering point, called sink (or base station), and communicate with each other through wireless communication links. In fact, they need to cooperate and collaborate with each other in order to accomplish the goals of their missions. As a consequence, the sensors should know their own locations as well as those of their neighbors. Therefore, knowing the sensors’ location is a crucial aspect for the design and development of this type of network. As a matter of fact, there are situations, where the sensors are deployed for bush fire surveillance, water quality monitoring, and precision agriculture. In these cases, the sensed data obtained by the sensors would be meaningless without knowing the location of where the data is captured. Each sensor node must be aware of its exact location before transmitting the cognate data. Knowing the location would open numerous possibilities to other applications, such as road traffic monitoring, intrusion detection, health monitoring, and surveillance (Mao, Fidan, & Anderson, 2007). Thus, different self-localization techniques arise in wireless sensor networks.

Localization is the process of finding the location of a sensor in a given coordinate system with the help of some special sensors, called reference nodes (Majhi, & Soni, 2014; Savvides, Han, & Strivastava, 2001; Priyantha, Miu, Balakrishnan, & Teller, 2001). This process serves to aid in network topology management, location identification of gathered data, evaluation of node density and coverage, geographical (or location-based) routing, object tracking, and other geographic algorithms. To this end, global positioning system (GPS) receivers can be used to determine the nodes location. Unfortunately, their usage incurs higher energy consumption and increases the cost of overall network. Therefore, it is essential that localization algorithms be energy-efficient and cost-effective. In this regard, various algorithms and protocols have been developed for sensor localization in both two-dimensional and three-dimensional wireless sensor networks. However, most of the established localization algorithms focus on two-dimensional sensor deployment. These algorithms are accurate on flat terrains. But, when they are used in harsh terrains, they produce less accurate results. Therefore, three-dimensional localization algorithms are required for better accuracy in these types of terrains. In this book chapter, we propose to study the sensor localization problem in three-dimensional wireless sensor networks. More precisely, this book chapter’s sole focus will be on three-dimensional sensor deployment, and it aims to provide an overview of the existing solutions to the localization problem in a three-dimensional space. Basically, it proposes a classification of localization algorithms, and discusses different three-dimensional sensor localization approaches along with their motivation and evaluation. This evaluation is based on different attributes, including accuracy, cost, computational complexity, and power consumption.