A Comprehensive Tutorial on Localization: Algorithms and Performance Analysis Tools

摘要

This tutorial offers a comprehensive view of technological solutions and theoretical fundamentals of localization and tracking (LT) systems for wireless networks. We start with a brief classification of the most common types of LT systems, e.g. active versus passive technologies, centralized versus distributed solutions and so forth. To continue, we categorize the LT techniques based on the elementary types of position-related information, namely, connectivity, angle, distance and power-profile. The attention is then turned to the difference between active and passive LT systems, highlighting the evolution of the localization techniques. Motivated by the interests of industry and academia on distance-based active localization system, a deep review of the most common algorithms used in these systems is provided. Non-Bayes-ian and Bayesian techniques will be tackled and compared with numerical simulations. To list some of the proposed approaches, we mention the multidimensional scaling (MDS), the semidefinite programming (SDP) and the Kalman filter (KF) methods. To conclude the tutorial, we address the fundamental limits of the accuracy of range-based positioning. Based on the unifying framework proposed by Abel, we derive the closed-form expressions for the Cramer-Rao lower bound (CRLB), the Battacharyya Bound (BB), the Hammersley-Chapmann-Robbins Bound (HCRB) and the Abel Hybrid Bound (AHB) in a source localization scenario. We show a comparison of the aforementioned bounds with respect to a Maximum-Likelihood estimator and explore the difference between random and regular (equi-spaced anchors) network topologies. Finally, extensions to cooperative scenarios are also discussed in connection with the concept of information-coupling existing in multitarget networks.