Particle Filter GNSS

Problem to solve and initial situation

Technological solutions for GNSS based navigation in weak signal situations are still far from being completed, still lacking the required accuracy and also reliability. Especially indoor navigation still gains rapidly increasing attention, considering the wealth of potential applications.

Particle filtering algorithms at the combined GNSS signal processing and positioning level may be considered as a very promising approach, however employment in operational receivers is not known yet. This technology is far from being mature and further investigation is needed.

Especially with Galileo becoming operative - presently in IOV phase and soon in FOC phase - a further push to such activities is being expected respectively already ongoing, as e.g. indicated by GNSS share in the Horizon 2020 Space budget.

Since many years JOANNEUM RESEARCH (JR) has been involved in GNSS research, mainly regarding propagation and interference issues. IFEN GmbH has excellent expertise in GNSS receiver architecture and signal processing, perfectly complementing JR's expertise.

IFEN GmbH is a German company with some of their employees living and working in Graz, close to JR premises, paying wage tax in Austria. JR and IFEN plan the implementation of particle filtering algorithms as a strategic important step, expecting superior performance in weak GNSS signal situations (indoors, urban canyons, railraod scenarios, but also space applications ...) and significantly enhancing JR's expertise.

Planned goals and approaches

A GNSS particle filter algorithm will be developed and implemented, and the performance of this implementation will experimentally be validated. The algorithms will be based on signal processing in position domain, direct use of correlator values, involvement of Bayesian filters and multipath handling.

The actual implementation work will be done by JR under guidance and supervision by IFEN. Particle filtering algorithms cannot be considered as stand-alone software solutions, instead always require certain aiding information, like precise time, or situational awareness (e.g. building layout, etc.).

The readily implemented algorithms will be validated by experiments, using indoor, urban canyon and railroad scenarios. Any shortcomings being revealed by the experiments will be overcome by improving the algorithm / implementation. The implementation will consider GPS and Galileo signals and may compare receiver performance considering just one of the two systems.

Expected results and findings

The expected result is a fully validated implementation of a GNSS particle filtering algorithm, ready for integration into a test bed or commercial product. Both, JR and IFEN, have full rights on the results, including the full source code.

Such algorithm and the experience gained in this activity will significantly enhance JR's competitiveness for upcoming opportunities, either in ESA's EGEP programme, under the relevant calls of GSA / Horizon2020, or in any other relevant business chances. IFEN may integrate the implementation into their commercial products as well.

Coordinator

JOANNEUM RESEARCH Forschungsgesellschaft mbH

Project partner

IGASPIN GmbH

JOANNEUM RESEARCH Forschungsgesellschaft mbH
Leonhardstrasse 59
A-8010 Graz
Web: www.joanneum.at