Indoor Multipath Characterization and Separation Using Distortions in GPS Receiver Correlation Peaks

V. Bellad, M.G. Petovello

Abstract:	Pseudorange measurements suffer from a variety of systematic biases and errors, most of which can be eliminated by differential positioning as they are correlated between the remote and base receiver. The measurement bias caused by signal multipath behaves differently because unlike other errors, multipath is normally uncorrelated between antenna locations. Hence, the base and remote receivers experience different multipath and as a result differencing between them will not cancel the errors. Multipath and ionospheric errors are the largest source of errors in single point positioning. However, ionospheric and other smaller ranging errors can be removed or reduced either by corrections from a base station or from SBAS satellites in standalone receivers. Multipath remains a major pseudorange measurement error source in both standalone and differential mode, especially in indoor and degraded signal environments. There are several methods for multipath mitigation at the receiver level. The most common methods of reducing multipath include antenna design to reject multipath signals arriving from ground and in some cases by careful site selection. Unfortunately, site selection is feasible only for very few specific applications and not much can be done for multipath signals arriving by reflections from surfaces above the horizon. The next point in the chain where multipath errors can be reduced is at signal processing level within the receiver. Various methods have been proposed for the reduction and estimation of multipath errors at the signal level in the receiver based on the chip spacing and other techniques. Prominent among them are Narrow correlator (Van Dierendonck et al 1992), Strobe correlator, Double delta correlator and Multipath Estimating DLL (MEDLL) (Townsend et al 1995). However, all these techniques are successful in reducing multipath errors under the assumption that the Line-Of-Sight (LOS) signal is stronger than the composite multipath signal. Although MEDLL estimates all the parameters causing the distortion in the correlation peak, it requires a bank of correlators and resources which become impractical for commercial applications. In the scenarios where multipath signal is stronger than the LOS signal these correlator techniques cannot identify multipath and continue tracking the multipath peak. To this end, it becomes important to be able to identify and separate LOS and stronger multipath signals in the correlator domain. Xie & Petovello (2011) proposed a block processing technique used in high sensitivity receivers to separate and identify LOS signals from Non-Line-Of-Sight (NLOS) signals for vehicular applications. This method exploits the fact that the Doppler for LOS and NLOS signals are different when the receiver is under dynamic condition. However, for static and low dynamic receivers with multipath peak falling within one chip or slightly more of the true peak, the presence of stronger composite multipath signals needs to be determined in the code correlation domain. Unless detected as tracking multipath peak, such channels can have significant pseudorange biases resulting in correspondingly large position errors. In this paper, the distortion in the correlation peak due to multipath will be investigated to understand the effects of multipath, and specifically, attempts will be made to separate LOS and multipath signals for static and low dynamic receivers. It is to be noted that the method so proposed shall be equally valid for higher dynamic applications. For a given front-end bandwidth, the shape of the ideal correlation triangle is effectively known. Distortions from this ideal case result from multipath and/or interference. Assuming that interference is identified elsewhere within the receiver, the focus here will be on identifying distortions resulting from multipath only. Given that multipath is, by definition, the result of signals arriving after the LOS signal, consideration will be given to how this might be exploited for the case at hand. Particularly, slope distortions on the early side of the correlation triangle will be assessed to detect the presence of a stronger NLOS signal. The performance evaluation of the above method will be initially done using simulated data generating multipath with different power and/or path delay values relative to the LOS signal. By determining the success of the algorithm in different circumstances, modifications will be made to improve the NLOS signal detection performance. However, like with any other techniques, the detection of NLOS signals through the investigation of correlation peaks is expected to have some limitations depending on the separation of LOS and NLOS peaks, relative power of LOS and NLOS signals, etc. The detection performance of the method will be evaluated using Receiver Operating Characteristic (ROC) probability curves as a function of these parameters. Real data from severe multipath environments will be used to verify the applicability of the method for static and different dynamic applications. The positioning accuracy performance will be evaluated and quantified by excluding or down-weighting erroneous measurements due to multipath detected through this method. References Van Dierendonck A.J., P. Fenton, and T. Ford (1992), “Theory and performance of narrow correlator spacing in a GPS receiver,” NAVIGATION, Journal of the Institute Of Navigation, 39(3) (Fall 1992), pp. 265-283 Xie P. and M.G. Petovello, and C. Basnayake (2011), “Multipath Signal Assessment in the High Sensitivity Receivers for Vehicular Applications,” Proceedings of ION GNSS-2011, 24th International Meeting, Portland, OR, September 2011, pp. 1764-1776 Townsend B., RDJ Van nee, P. Fenton, and A.J. Van Dierendonck (1995), “Performance evaluation of the multipath estimating delay lock loop,” ION National Technical Meeting, Anaheim, California, January 18-20, 1995
Published in:	Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013) September 16 - 20, 2013 Nashville Convention Center, Nashville, Tennessee Nashville, TN
Pages:	3564 - 3576
Cite this article:	Bellad, V., Petovello, M.G., "Indoor Multipath Characterization and Separation Using Distortions in GPS Receiver Correlation Peaks," Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013), Nashville, TN, September 2013, pp. 3564-3576.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In