Towards dynamic array processing for GNSS software receivers

Sammanfattning: Since the very first Global Positioning System satellite was launched in 1978, Global Navigation Satellite Systems (GNSS) have developed into a world wide utility, providing everyone with an affordable method for determining accurate position and time. However, the system is susceptible to multipath (when the signal is reflected off a surface), interference (other signals in the same frequency band) and attenuation (for example canopies blocking some of the signal energy). Array processing of GNSS signals have lately drawn quite a bit of attention from the research community, where issues (susceptibility to interference, multipath and attenuation) can be mitiged to some degree. GNSS baseband processing is generally implemented in hardware (digital logic), although software based processing are swiftly gaining popularity among researchers. This thesis discusses the implementation of a software based GNSS array processing system with an emphasis on hardware, calibration and initial signal detection. Low cost, ASIC based front-ends are examined with regards to phase and gain stability, and are found to meet the requirements for array processing of GNSS signals. An antenna calibration technique using live GNSS signals (as opposed to anechoic chamber measurements) is proposed, where the predictable orbits and geometric diversity of GNSS are exploited. A system capable of replaying recorded narrow band GNSS signals into any low cost receiver is also discussed. Finally, efficient methods to acquire GNSS signals using an antenna array system are proposed.