Article Preview
Top1. Introduction
Wireless Sensor Networks (WSNs) are a popular field of research as there are many aspects to improve upon. A WSN normally consists of architecturally primitive sensors that can have many different applications like landslide detection, water quality monitoring, forest fire and many more. It is important in such applications to know the location where from the data is coming, in order to take necessary required actions. However, the required precision of the location information depends on the application, and quality and capability of the sensor node.
In general, WSNs are composed of the sensors which do not have any mechanism of finding their own location. While all the sensors have wireless connectivity, it is possible to acquire some location information from neighboring nodes. However, it is communication which drains the battery most in WSN (Alippi 2010), therefore it is important to use the power/energy efficiently while acquiring location information in order to increase the lifetime of network. To keep production cost low, sensors are typically not equipped with self-localization equipment such as Global Positioning Systems (GPS) and must calculate their location based on other sensors. Sensor nodes that know their own location are referred as anchors. Self-localization in WSNs can be done by using techniques such as triangulation, trilateration and Maximum Likelihood estimation (Amitangshu, 2010). These make use of ranging techniques such as Angle-Of-Arrival (AOA), Time-Of-Flight (TOF), Time-Difference-Of-Arrival (TDOA), and Received Signal Strength Indicator (RSSI). All the ranging techniques are prone to errors in precision (Mao 2007; Tanvir 2010). Some of the techniques are more precise than others, but this precision requires more expensive equipment and these techniques are therefore not feasible in commonly used WSNs.
Figure 1. Example localization wave in wireless sensor network
In this paper the authors focus on multi-lateration using Two-way ranging technique based on time of flight. The multi-lateration is when four or more anchor nodes are used to calculate location. To self-localize a sensor node at least requires three anchors, which is generally referred as tri-lateration. A trilateration or multi-lateration uses the distance measurements between anchor nodes and the un-localized nodes estimated based on two-way ranging technique. A detailed description about multi-lateration is given in Section 5. One problem with Two-way TOF is that the distance measurements can be affected by noise such as weather, temperature and physical obstacles (Thelen, Goense, & Langendoen, 2005). When using tri/multi-lateration, the error gets accumulate throughout the network resulting in increased inaccuracy the farther away a sensor is from the original anchors (Huang 2009). The WSN topology similar to one shown in Figure 1 is used while conducting the experiments. The main focus is to minimize error accumulation throughout such topologies. Three anchor nodes are placed at the center of the network, and then using these three anchor nodes and other localized nodes, other nodes (un-localized) are localized.