Signal processing and collective detection for locata positioning system

Abstract

Unlike conventional Pseudolite Positioning Systems (PLPS), Locata is a positioning technology that employs a network of terrestrial-based time-synchronised transceivers (LocataLites) that transmit pulsed signals (i.e. Time-Hopped Direct Sequence Code Division Multiple Access or TH/DS-CDMA signal) in a non-GNSS frequency band, effectively mitigating most of the problems faced by conventional PLPS. However, the signal processing methods of TH/DS-CDMA are largely unreported in the academic community. This thesis focuses on the signal processing techniques (i.e. acquisition and tracking) for Locata and the possibility of integration with GPS at the signal acquisition stage. The TH/DS-CDMA signal was fully characterised and its signal parameters identified without any knowledge of its Interface Control Document (ICD) or similar documents. The full characterisation allowed investigations on the acquisition and tracking of Locata s signal to be conducted from the ground up . Assuming sufficiently strong signals, two exhaustive search methods that attempt to minimise acquisition time are proposed to synchronise the TH component at the acquisition stage, an area of study that has not been previously addressed. Furthermore, the TH acquisition process is extended to combine signals of multiple bursts to overcome weak signals via the Sequential Probability Ratio Test (SPRT) concept. A method to track the TH/DS-CDMA signal is proposed and compared against benchmark cases. A modified tracking loop is then used to track combined signals via a simplified structure using fewer components, and proved to outperform a single channel tracking loop. The performance of various estimation approaches for a recently developed multi-transmitter combined acquisition method (Collective Detection) for GPS L1 signals was evaluated. The method was empirically proven under a typical scenario to be more sensitive than conventional acquisition methods by a margin of 3dB. The method was expanded to include terrestrial Locata signals, making possible the first investigation of GPS/Locata integration at the signal acquisition stage. Collective Detection was also evaluated for its performance under various coarse-time errors, a type of error that has not been previously accounted for. A hybrid Collective Detection algorithm was proposed and shown to reduce significantly the computational complexity of the original approach by accounting for incompletely detected pseudoranges.