ION GNSS 2012
Session F1: Urban & Indoor Navigation: GNSS & Assisted-GNSS

Title: Investigating Indoor GPS Doppler and Pseudorange Characteristics
Author(s): S.N. Sadrieh, A. Broumandan and G. Lachapelle, University of Calgary, Canada
Date/Time: Wednesday, September 19, 2012, 11:48 a.m.
Room: Grand Ballroom Center (Renaissance)

Although GPS performs well in open sky circumstances, in harsh multipath environments such as indoor locations it suffers degraded performance to provide accurate and reliable position. It is due to the fact that GPS signal propagation in these environments is a complex phenomenon wherein many of the physical effects such as reflection, diffraction, attenuation and scattering affect the Radio Frequency (RF) wave received by an antenna. Hence, the received signal is superposition of multiple copies of a unique transmitted signal experienced different attenuations, delays and phase shifts. It gives raise to two main difficulties for a GPS receiver to operate in an indoor environment namely: signal attenuation and multipath propagation. In order to compensate for these shortcomings, advanced methods of weak signal detection and parameter estimation and integration of GPS with inertial sensors (GPS/INS) can be introduced as appropriate approaches for indoor navigation. Performance of these approaches highly depends on different characters of propagation channels. GPS/INS integrated systems utilize pseudorange and Doppler measurements to update the INS drift. In such systems, the accuracy of GPS measurements plays a critical rule for tuning the integration filter (Petovello 2003). Applicability and performance of advanced weak signal detection and parameter estimation depend on channel characters. For example, time diversity methods relay on independency of received signal over time whenever increasing the coherent integration time is applicable when the received signal is correlated over time (Broumandan et al 2011). Thus, it is of interest to characterize the indoor GPS channel more specifically in term of indoor Doppler and pseudorange accuracies. Independency or correlation of measurements error over time or space can affect the signal processing strategies. Ray et al (2001) assumed that for outdoor scenarios, the error due to multipath is spatially correlated. Hence, they have used spatially separated antennas to estimate the error and make some corrections. However, for indoor circumstances, this correlation is seriously questioned by the results shown in Sadrieh et al (2010). Several efforts have been made to characterize the indoor measurement error sources. The receiver thermal noise typically induces an error on Doppler and pseudorange measurements. These errors are functions of Signal-to-Noise Ratio (SNR). Thus, in signal-degraded environments, the error due to noise will be more considerable. In Stein (1981) and Kay (1998) the Cramer-Rao Lower Band (CRLB) of Time-of-Arrival (TOA) and Frequency of Arrival (FOA) are presented as functions of SNR and spectrum of transmitted signal. Considering the GPS spectrum, receiver bandwidth and filter these bounds limits the pseudorange and Doppler measurements accuracies. Borio et al (2011) reported that Doppler estimation accuracy under attenuated circumstances is close to CRLB. In Sadrieh et al (2012) Power Spectral Density (PSD) of the received GPS signal is formulated and verified using indoor live GPS signal for different multipath models and antenna gain patterns. The Effect of multipath models and antenna gain patterns on the accuracy of Doppler measurements were also investigated . In Jost & Wang (2009), the Power delay profile of a satellite to indoor channel is assessed in C band. However, the significant difference between bandwidth, signal power and spectral shape of transmitted signal and actual GPS signal make the comparison hard. This research conducted to investigate the accuracy of indoor measurements in different places and characterize the error sources. The metrics to be investigated are; Doppler and pseudorange error due to attenuation and multipath, power spectral density and power delay profile. In order to achieve the above goals, raw GPS L1 signal from three antennas are captured simultaneously; one reference outdoor antenna, one outdoor attenuated signal, and one indoor moving antenna, which emulates the user motion in indoor environments. Using block processing method and increasing the coherent integration time enables the receiver to acquire GPS signal indoors. In order to characterize indoor error source, other common error sources are removed by Differential GPS (DGPS) processing. Herein two DGPS measurements are assessed. The first one is established between the reference antenna and attenuated channel. The second DGPS is established between the reference antenna and an indoor moving antenna. In both cases, orbital and atmospheric errors are eliminated due to very short baseline. The first DGPS measurements are effected by noise only where the second DGPS measurements experienced noise and multipath. Jointly analyze the accuracies of both DGPS measurements yields to a comprehensive understanding of noise and multipath effects on indoor measurements. More insight is achieved by comparing the achieved accuracies with the CRLB. The initial results show that Doppler measurement is more corrupted by multipath rather than attenuation. Multipath is the main source of error on pseudorange measurements in mid-attenuated circumstances (e.g. wooden houses) where in extremely attenuated circumstance the noise effect is dominant.
References: Borio D., N. Sokolova and G. Lachapelle (2011) "Doppler Measurement Accuracy in Standard and High-Sensitivity GNSS Receivers," IET Radar, Sonar & Navigation, vol 5, no 6, pp 657-665. Broumandan, A., J. Nielsen, and G. Lachapelle (2011) "Coherent Integration Time Limit of a Mobile Receiver for Indoor GNSS Applications," GPS Solutions, Published online March 2011, 11 pages

Jost, T. and W. Wang (2009) "Satellite-to-indoor broadband channel measurements at 1.51 GHz," in Proceedings of the ION GNSS International Technical Meeting, 26-28 Jan , Anaheim, CA, pp. 777-78

Kay, S. M. (1998) Fundamentals of Statistical Signal Processing: Detection Theory, Prentice-Hall.

Petovello, M. G. (2003) Real-time Integration of a Tactical-Grade IMU and GPS for High-Accuracy Positioning and Navigation, PhD Thesis, Department of Geomatics Engineering, University of Calgary, Canada Ray, J.K., M.E. Cannon and P. Fenton (2001) "Code and Carrier Multipath Mitigation Using a Multi-Antenna System", IEEE Transactions on Aerospace and Electronic Systems, vol 37, no 1, pp.183-195. Sadrieh, N., A. Broumandan and G. Lachapelle (2010) "Spatial/Temporal Characterization of the GNSS Multipath Fading Channels". In Proceedings of GNSS10, The Institute of Navigation, Portland OR, 21-24 September 2010. Sadrieh, N., A. Broumandan and G. Lachapelle (2012) "Doppler Characterization of a Mobile GNSS Receiver in Multipath Fading Channels" .The Journal of Navigation, Cambridge University press, In press Stein, S. (1981) "Algorithms for Ambiguity Function Processing". IEEE Transactions on Acoustics, Speech and Signal Processing, vol 29, no 3, June 1981.

Previous Abstract Return to Session F1 Next Abstract