Scheme for enhanced tracking of mobile subscribers in a WiMAX network

Abstract

In this paper, the base station identification and timing adjust measurements are used to geolocate mobile subscribers in a WiMAX network. The uplink and downlink subframes of the physical layer and management messages of the medium access control layer are examined to extract the necessary data for geolocation. Using a hidden Markov model [1] based algorithm to estimate the track of the mobile subscriber, we demonstrate that the position error can be further reduced by incorporating timing adjust measurements. Simulation results of the proposed scheme are included to demonstrate the effectiveness of the combined use of base station ID and timing adjust measurements.

A computationally efficient approach for hidden-Markov model-augmented fingerprint-based positioning

“…As our study focuses on network deployment using modern 4G standards, we find it convenient to exploit the built in TA in conjunction with associated base station IDs (BSID) to this end. Simulation has shown that TA data can dramatically improve accuracy when such data are available (Henderson, 2011;Henderson et al, 2012). When it is not available, or if it is not available at a sufficiently frequent interval, the proposed HMM-augmented fingerprinting method described in the preceding section can be used unassisted.…”

“…Scalability is the central theme of our contribution enabling significant reduction in computational cost. Additionally, where other fingerprinting localisation schemes typically use only a two-phase program (offline database training and online localisation) (Bshara et al, 2011;Henderson et al, 2012;Malik & Allen, 2006;Nypan et al, 2002;Roxin et al, 2007;Xu & Jiang, 2012), our proposed scheme uses three phases: an offline database training phase, an online International Journal of Systems Science 3 localisation phase, and optional HMM and TA augmentation. Our technique of TA integration furthers our ability to scale the data without compromising accuracy.…”

“…This signal of interest may be the channel impulse 2 J. Roth et al response (Kabir & Kohno, 2012;Malik & Allen, 2006), received signal strength (RSS) (Bshara, Orguner, Gustafsson, & Van Biesen, 2010;Feng, Au, Valaee, & Tan, 2010), or cell-ID measurements (Bshara, Orguner, Gustafsson, & Van Biesen, 2011;Henderson, Tummala, McEachen, & Scrofani, 2012;Xu & Jiang, 2012). During the tracking phase, this signal is then observed and compared to the database measurements via some measure of similarity.…”

“…However, where Bshara et al (2011) assumed that a motion model always increases accuracy, we show that this is not true and further discuss the scenarios where their assumptions do not hold. Computational sensitivity is achieved through introducing a scalable approach (Henderson, 2011;Henderson et al, 2012) to well-established techniques. This paper formalises computational concepts formerly treated heuristically (Henderson et al, 2012) and quantifies the actual computational gains achievable by this novel approach.…”

“…Computational sensitivity is achieved through introducing a scalable approach (Henderson, 2011;Henderson et al, 2012) to well-established techniques. This paper formalises computational concepts formerly treated heuristically (Henderson et al, 2012) and quantifies the actual computational gains achievable by this novel approach. Furthermore, through building on the work begun by Whitty, Tummala, and McEachen (2012) and Barber (2009), a second novelty is realised enabling even more drastic data scaling by integrating timing adjust (TA) as a third option for augmentation.…”