WHLocator: hybrid indoor positioning system

Reviewer: Arun Ektare

Global positioning systems (GPSs) have become very popular because of their ability to locate an entity with accuracy. GPS uses satellites to find the position of a location. Indoor positioning systems use WiFi, a technology widely adopted in indoor environments. This paper addresses the need for inexpensive indoor positioning systems based on WiFi technology. It describes WHLocator, a hybrid indoor positioning system. The paper consists of nine sections. In Section 1, "Introduction," Lemieux and Lutfiyya describe the methodology of locating an object indoors. For preselected locations, the received signal strength indicator (RSSI) values are known. By comparing the RSSI of the unknown location to those known, the unknown location can be established. However, the RSSIs are subject to fluctuations created by movement, reflection, noise, and type of hardware used. This section discusses several scenarios where these fluctuations make it difficult to estimate the position of a location. Section 2 describes the major approaches pursued in this area, including specialized hardware that uses radio frequencies, ultrasound, or both. The authors discuss some available applications and describe a grid system called RADAR, which uses signal strengths at grid intersections to locate an object. Another system, called Horus, uses signal strength probability for locating an object. Finally, hybrid systems are described that use infrared technology with stereo vision trackers and inertial navigation systems (INSs); these are still under development. Section 3 "discusses the algorithms used by WHLocator." It presents the pseudocode and offers detailed descriptions. Next, it describes the use of altimeter readings to filter out the locations reported that are on the wrong floor. Then, it describes a scene classifier that calculates the probability that a scene is a particular type-for example, classroom, computer lab, or office. Finally, this section describes the sequence that is followed to find a location. Section 4, "Implementation," describes the hardware, software, altimeter, and camera. There is also a discussion of the software developed using Java. Section 5 is a detailed discussion of an experimental setup. The experiment was carried out in a college that has multiple-floor buildings made up of classrooms, computer labs, and offices. This section briefly covers fingerprint data collection, the object's locations, and scene analysis involving user movement. Section 6 presents the performance results. Section 7 concludes with a discussion of the results, including the benefits of altimeter and scene analysis and possible improvements to the approach. Sections 8 and 9 present the authors' acknowledgments and a list of references, respectively. While the paper is well written, the diagrams are rather difficult to read. Despite this, the paper may be useful to practitioners. Online Computing Reviews Service