Localization Algorithms and Strategies for Wireless Sensor Networks: Monitoring and Surveillance Techniques for Target Tracking

Introduction

Wireless sensor networks are ideal candidates for data gathering and remote sensing purposes in various environments, where the communications between the sensors mostly take place in a distributed manner (Akyildiz et al., 2002). Data obtained by individual sensors are relayed to a central station for further processing and logging. Usually, information obtained through a sensor needs to be associated with the location of each sensor which necessitates the localization of sensors within certain accuracy (Patwari et al., 2005). Since primary purpose of sensor networks is to gather information on environmental changes such as temperature, pressure or humidity, it is almost always required to determine the coordinates of a specific sensor so that the appropriate steps can be taken in a more effective way in the case of emergencies. As a matter of fact, about 13.3% of the recent scientific WSN publications focus on target tracking and localization aspects as mentioned by Sohraby et al. (2007).

Owing to their small form factor and low-power consumption, WSNs have found numerous applications in both civil and military use. In the civil domain, applications can be further divided into various categories like environmental, health related, commercial and public safety applications. Environmental applications may include fire/flood detection/prevention (Pathan et al., 2006), crop quality detection, field surveying; health related applications may be listed as patient/doctor/instrument tracking inside hospitals, elderly care and remote monitoring of biological data (Cypher et al., 2006); commercial applications might be inventory control, product tracking in warehouses (Rohrig & Spieker, 2008) and remote product quality assessment. For military applications, the use of WSNs is also important in rough terrain conditions where a centralized communication system may be too costly to build (Merrill et al., 2003). Soldier and mine tracking, as well as intelligence gathering can be some applications in this domain.

As it can be seen from a sample of applications in each field, location and tracking capability is an important aspect of WSNs that need to be considered and developed further. A number of researchers studied positioning using sensor nodes and investigated the performance of algorithms and presented theoretical bounds in WSNs (Bulusu et al., 2000; Niculescu et al., 2001; Savarese et al., 2001; Savvides et al., 2001; Doherty et al., 2001; Chang & Sahai, 2004; Kanaan et al., (2006a, 2006c)). The positioning capability can be implemented using different sensing technologies like ultrasonic waves as in the Active Bat system proposed by Ward et al. (1997) or using RF or both as in the Cricket system (Priyantha et al., 2000). By using sound waves, researchers have been able to obtain cm. accuracy; however, these systems are only suitable for very small areas like a single office environment and are not intended for outdoor or indoor/outdoor hybrid node localization. In the latter case, RF solutions are generally preferred since quick deployment is possible and hardware and various ranging/localization algorithms that can be directly applied are widely available. Nevertheless, due to the nature of RF propagation, ranging/localization accuracy is not on par with solutions using sound waves. Later in this chapter we will cover the basics of RF channel and how it affects the performance of localization in various environments.

Figure 1, shows a typical setup for a WSN with location capability. Here, the sensor nodes are able to communicate with each other as well as pre-deployed anchor nodes whose coordinates are known in advance. At this point, it might be appropriate to present the two methods of WSN localization. Sensor node localization can take place in a centralized or a distributed manner.