A Taxonomy of Localization Schemes for Wireless Sensor Networks A. Youssef Google Inc. Mountain View, CA, USA M. Youssef Computer Science Department University of Maryland College Park, MD, USA Abstract Knowledge of nodes’ locations is an essential re- quirement for many applications. This paper surveys the cur- rent state of the art for localization schemes in sensor net- works. We present a taxonomy of the localization schemes for sensor networks based on different features. We then describe how the current localization schemes for sensor networks map to these different features. We believe that this paper serves as an introduction for researchers interested in the area of local- ization schemes for sensor networks as well as in evaluating the characteristics of a location system needed by a particular application or the suitability of an existing location system for the application. Keywords: node localization, scalability, ad-hoc networks, wireless sensor networks, algorithms. 1 Introduction Location discovery in sensor networks have been an active research area for the past couple of years. Ap- plications for localization schemes for sensor networks include determining the location of an event, location- based routing, e.g. [1], node identification, and node coverage, e.g. [2]. Several localization algorithms have been proposed and implemented. In this paper, we describe a taxon- omy that classify the localization schemes for sensor networks based on several distinct features. The tax- onomy can be used to understand the key features that distinguish different localization schemes and help in selecting the appropriate scheme for a particular appli- cation. In addition it introduces new comers to the area of localization schemes for sensor networks. Based on this taxonomy, we present an overview of the current research in this field of localization schemes for sensor networks by surveying several localization algorithms. This survey helps in understanding the dif- ferent features of the taxonomy as well as understanding the current state of research in this field. The rest of the paper is organized as follows. In Section 2 we present our taxonomy of the localization schemes for sensor networks. Section 3 surveys the cur- rent work in the filed of localization schemes for sensor networks. We conclude the paper in Section 4 and give directions for current trends in the area of localization schemes for sensor networks. 2 Taxonomy Features Location discovery algorithms may be classified ac- cording to several criteria, reflecting fundamental de- sign and implementation choices. Those different crite- ria form a reasonable taxonomy for characterizing and evaluating location discovery algorithms. In this sec- tion, we summarize different design alternatives for lo- cation discovery algorithms in general and in wireless sensor networks in particular. Figure 1 shows our pro- posed taxonomy. The rest of this section describes the taxonomy in more details. 2.1 Anchor-based versus Anchor-free Anchor-based algorithms operate on an ad-hoc net- work of sensor nodes where a small percentage of the nodes (anchors) are aware of their positions ei- ther through manual configuration or using GPS. An- chor nodes broadcast their locations information to their neighbors. The goal is to estimate the positions of as many unknown nodes as possible using anchor node in- formation. Anchor-based algorithms usually produce an absolute location system where absolute node posi- tion is known, for example, latitude, longitude, and al- titude. However, the accuracy of the estimated position is highly affected by the number of anchor nodes and their distribution in the sensor field [3]. Langendoen et al. [4] showed that with anchor density of 20%, we could have an accuracy of 25% of transmission range, which falls short from the required inaccuracy in many