Article,

Tracking Complex Modulation Waveforms - How to avoid receiver bias

, , , and .
Presentation at the Position Location and Navigation Symposium (PLANS), (2006)

Abstract

Analysis of the open service modulation waveforms proposed for the Galileo signal on L1 shows that the combined binary coded symbol (BCS) and BOC(1,1) waveforms, as proposed at GNSS ION 2005, may induce tracking bias in a receiver designed to receive only the BOC(1,1) waveform. Navigation using a mixed constellation of modernised GPS satellite broadcasting a BOC(1,1) component and Galileo with one of the possible additive mixtures of BCS and BOC(1,1) - would lead to potential location errors if no corrective measures are taken. The bias depends upon the correlator spacing in the code tracking detector. There are several ways to calibrate the bias. One of these uses a stored table of corrections for use with Galileo signals whose entries depend upon the receiver parameters. Another method uses additional satellite signals and establishes a filter state in a Kalman navigation processor. In this case, the presence of multi-path propagation makes such calibration difficult for users with a finite observation time. An alternative solution uses phase alternation of the BCS component at the chip rate to eliminate the measurement bias for correlator spacings smaller than 0.4 chips. The trend is indeed to use smaller and smaller correlator spacings in the near future. The paper illustrates the tracking bias with examples from the current Galileo signal proposals. The paper establishes an effective method of removing or avoiding the bias corresponding to changes in the satellite signal. We show that the tracking bias is sensitive to the choice of the BCS waveform and to its use in a receiver adapted for BOC(1,1) reception. The main contribution of the paper is to show that there are BCS sequences which lead to bias-free pseudo-range estimates through appropriate design rules. These belong to a set known as anti-symmetric sequences. The resulting set of allowed BCS sequences is very limited in number. There are none of sequence length 10, as proposed originally for Galileo. The spectral properties of the best sequence is shown. There are consequential changes to the performance of the composite spreading symbol sequences in the presence of multi-path propagation and these effects are also shown and may be compared to previous Galileo signal proposals.

Tags

Users

  • @bmuth

Comments and Reviews