ION GNSS 2012
Session D3: GNSS Algorithms & Methods 1: Signal Processing

Title: Assisted GNSS in LTE-Advanced Networks and its Application to Vector Tracking Loops
Author(s): C. Fernandez-Prades, P. Closas, Centre Tecnologic de Telecomunicacions de Catalunya, CTTC, Spain; J Vila-Valls, M. Figueroa Naharro, A. Torne, Sanjose, Universitat Politecnica de Catalunya, Spain
Date/Time: Thursday, September 20, 2012, 11:48 a.m.
Room: 204 (NCC)

This paper describes the positioning protocols provided by Long Term Evolution (LTE) networks [Spi12], making special emphasis on existing and forthcoming standards for Assisted Global Navigation Satellite Systems (A-GNSS) data, mainly aimed to shorten the Time To First Fix (TTFF). The novelty of this work consists in proposing the usage of such information to enhance the performance of Vector Tracking Loops (VTL) in the user´s mobile terminal. Realistic simulations show improvements in terms of coverage, availability and accuracy with respect to receivers that also implement VTL but do not explicitly exploit A-GNSS data, or use it only to accelerate acquisition.
With conventional standalone GNSS, the GNSS receiver embedded in the mobile device is the sole responsible for receiving satellite signals and computing its location. It needs to acquire satellite signals through a search process, and it must lock tracking onto at least four satellites in order to compute a 3-D position. The acquisition process can be demanding in terms of battery and processing power, and TTFF can be long (up to 5 minutes in the worst case). The information provided by A-GNSS allows the purposely-equipped receiver to speed up the position calculation process, improving its sensitivity and helping to save battery power consumption. Currently, 3rd Generation Partnership Project (3GPP) Radio Resource Location Services Protocol (RRLP) for GSM networks is the standard de facto for providing GNSS assistance data to A-GNSS enabled mobile devices such as modern smartphones and tablets.

Cellular wireless technology is experiencing a rapid migration to a new generation of mobile broadband technology commonly known as Long Term Evolution (LTE) networks or, more specifically, "LTE-Advanced" or simply "4G" starting from 3GPP Specification Release 10, which was frozen in April 2011 (Release 11 is scheduled to be frozen by the end of 2012). As a matter of fact, LTE is enjoying strong industry commitment: according to Qualcomm [Qua12], 49 LTE networks are already deployed worldwide and 226 more will be available in the short term, providing service to a growing device ecosystem that currently is already formed by more than 270 different terminals of 60 different vendors. In addition of constituting the next performance leap in network capacity and wireless data throughput, this technology also defines protocols for user terminal positioning.

LTE positioning standards support three independent handset based positioning techniques: Enhanced Cell ID (ECID), Observed Time Difference of Arrival (OTDOA), and A-GNSS. After a brief review of each technique, we focus on A-GNSS and the related protocols: LTE Positioning Protocol (LPP), the Secure User Plane Location (SUPL) 2.0 and 3.0 and the Open Mobile Alliance (OMA) LPP Extensions (LPPe), which broaden the system´s functionality in the User Plane. These protocols are designed to provide assistance data to the user terminal (local atmospheric models, an approximate location coming from the serving cell tower, an approximate time accurate to a few seconds, a description of the satellite orbits and clock errors, and so on) for GPS, GLONASS, Galileo and QZSS.

The objective of this work is to enhance receivers´ performance by means of the exploitation of those assistance data in the tracking process, complementarily to their normal usage during signal acquisition. Specifically, we propose to include such data in the framework of VTL [Pet09], a signal processing technique consisting in jointly estimating the synchronization parameters of more than one satellite accounting for the geometrical dependence among them. The VTL is proven to deliver robust estimates in challenging scenarios, in contrast to traditional scalar tracking loops., which is proven to deliver estimations with lower variance than the estimates obtained by traditional scalar tracking loops.

Our proposed implementation of Delay/Frequency Vector Tracking Loop can incorporate approximate synchronization values of satellites that are above the horizon but are not in line-of-sight (or in other terms, with a receiving power below the receiver´s sensitivity). Since vector loop algorithms are based in the Bayesian estimation framework (that is, running a Kalman-like filter), it is quite natural to inject that "aprioristic" information, along with its estimated covariance, and obtain filtered position results.

VTL are known to bring performance improvements which are contingent on the number of available satellites and their geometry. By incorporating properly weighted A-GNSS information provided by LTE networks, we show that the resulting algorithm takes advantage of the augmented data set and outperforms state-of-the-art approaches in terms of availability, sensitivity and delivered position accuracy.

Computer simulations assess performance improvements with respect to traditional Delay/Phase Lock Loops and "non-assisted" VTLs, taking into account the specific format and characteristics of A-GNSS data provided by LTE networks in realistic, pertinent scenarios such as rural, sub-urban and urban environments. GNSS signal generation and propagation channel models were simulated in a high level of detail, following legitimate standards along with relevant propagation models such as the DLR Land Mobile channel [Clo12], and taking into account typical electrical parameters of commercial-off-the-shelf GNSS receivers.

In conclusion, this paper provides an overview of LTE-based positioning, with up-to-date information about the standardization status and emphasizing the A-GNSS service, and presents a new approach to take advantage of the infrastructure of the oncoming mobile cellular network technology, for which a massive deployment is envisaged in the short term. After providing mathematical details and the full description of the proposed algorithm, realistic computer simulations assess the gain of incorporating A-GNSS data in a VTL, showing benefits in term of position availability and accuracy, as well as in enhanced sensibility. All this reverts into an improved user experience by exploiting information that already exists in the receiver, while implying a modest computational complexity increase with respect to existing vector loop architectures.

References

[Clo12] P. Closas, C. Fern ndez-Prades, J. Diez, D. de Castro, "Multipath Estimating Tracking Loops in Advanced GNSS Receivers with Particle Filtering," IEEE Aerospace Conference, Big Sky (Montana) Mar. 2012.

[Spi12] Spirent White Paper, "An overview of LTE positioning", Sunnyvale, CA. February 2012.

[Qua12] Qualcomm, LTE Advanced. February 2012.

http://www.qualcomm.com/media/documents/lte-advanced-global-4g-solution

[Pet09] M. Petovello, M. Lashley, D.M. Bevly, "What are vector tracking loops, and what are their benefits and drawbacks?," Inside GNSS, May/June 2009, pp. 16-21.

Previous Abstract Return to Session D3 Next Abstract