A Robust Technique for Unambiguous BOC Tracking

J. Wendel, F. M. Schubert, S. Hager

Abstract:	The Galileo PRS signals, the Galileo open service signal on E1, and the GPS M code signals use a binary offset carrier (BOC) modulation. A BOC signal is obtained by modulating the carrier with the pseudorandom noise (PRN) code chips and a rectangular square wave subcarrier. Depending on the phasing of the square wave subcarrier w.r.t. the PRN code chips, a BOC cosine or a BOC sine signal results. Due to the modulation with the subcarrier, the signal energy is shifted to frequencies below and above the center frequency, while the frequency offset from the center frequency is given by the subcarrier rate. The autocorrelation function (ACF) of a BOC signal shows multiple peaks, while the envelope of the ACF is given by the triangle shaped correlation function of the PRN code. Therefore, the main peak of the ACF is sharper than the peak of the corresponding binary phase shift keying (BPSK) signal, which allows for a more accurate tracking, i.e. less tracking jitter, and an in creased multipath robustness. On the other hand, there is a risk that a tracking loop locks to a secondary peak, which results in a systematic error in the pseudorange measurements if this is not detected and countered. For example, this systematic error is a multiple of approximately 10m for the Galileo BOCc(15,2.5) signal on E1, and approximately 15m for the Galileo BOCc(10,5) signal on E6. Therefore, such a false lock to a sidepeak leads to errors in the position, velocity and time (PVT) solution and needs to be avoided. A variety of techniques exist for tracking BOC signals. A classical technique is bump-jumping. Hereby, additional correlators are used in order to assess the amplitudes of the peaks besides the peak to which the tracking loop is locked. Ideally, as the main peak of the correlation function is the largest, this allows for the detection of false locks. However, especially in the presence of multipath this technique is known to be unreliable. Another group of techniques reconstruct the triangle correlation function of the corresponding BPSK signal, for example by correlating with cosine and sine subcarriers or by a BPSK-like tracking of the upper and lower sidelobes. Obvisouly, this solves the ambiguity issue, but does not allow to exploit the potential of the BOC signal regarding increased multipath robustness and reduced tracking jitter. Another approach is the Double Estimator, where three independent but cooperative tracking loops are used: a phase lock loop (PLL) for the carrier, a delay lock loop (DLL) for the code, and a subcarrier lock loop (SLL) for tracking the subcarrier component. Both DLL and SLL produce delay estimates; the DLL delay estimate is unambiguous but less accurate, the SLL delay estimate is ambiguous and more accurate. The final pseudorange is obtained by a rounding operation, where the DLL delay estimate is used to fix the SLL delay ambiguity. This overview on BOC tracking techniques is by far not complete. A huge number of other approaches for solving the BOC tracking problem have been studied during the last years, which differ in performance, robustness, and hardware complexity. In this paper, an alternative approach is proposed which can be applied for any tracking technique providing an ambiguous delay estimate based on the subcarrier, and an unambiguous delay estimate based on the code, like the Double Estimator for example. The major difference to other techniques consists in the fact that the subcarrier ambiguity is not resolved for each tracking loop independently, but via position domain. This approach is based on the observation that the subcarrier ambiguity problem is mathematically identical to fixing carrier phase ambiguities, like it is done in real-time kinematics (RTK). Fixing carrier phase ambiguities requires that double-differenced measurements are used, measurements from a base station at a known position are mandatory. However, to fix subcarrier ambiguities is a much less demanding than fixing carrier phase ambiguities: The wavelength of the subcarrier is around 10 meters or more depending on the signal, while the carrier wavelength is around 20 cm. Therefore, for subcarrier ambiguity fixing a base station is not required, reliable results in stand-alone operation are perfectly possible. The basic idea is to calculate a PVT and float subcarrier ambiguities from the DLL and SLL delay estimates. Then, a state-of-the art technique for carrier phase ambiguity estimation like the LAMBDA method is used to calculate fixed subcarrier ambiguities, which are used to correct the PVT solution. Additionally, false locks are detected this way. In this paper, the proposed technique is described in detail, and simulation results are given which compare this technique to the Double Estimator, showing an increased robustness of the proposed approach. Finally, tracking structures different from the Double estimator are described for which this approach is applicable, too.
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:	3536 - 3547
Cite this article:	Updated citation: Published in NAVIGATION: Journal of the Institute of Navigation
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In