Performance of Antenna Array Calibration in Multipath Environments

Abstract

Antenna array processing is gaining attention in the GNSS community due to its ability to mitigate different types of interference. Array calibration is the first step towards implementation of most of the distortionless beamforming techniques. In the case of a GNSS, antenna array calibration can be performed using live signals. Since several satellite signals are available from different directions at the same time, the calibration process can incorporate all available GNSS signals. It is required to have a data set without any electronic interference or multipath to avoid errors in the calibration process. However, multipath is a major challenge for calibration and exists almost everywhere. The carrier phase measurements used to estimate the calibration parameters are affected by multipath, which introduces significant gain and phase errors. Due to this, the calibration parameters estimated would normally be erroneous. Carrier phase multipath in GNSS is periodic in nature with zero mean. This feature is exploited herein to improve calibration in multipath environments. By filtering the relative phase measurements between the antenna elements over several multipath periods, estimation of calibration parameters can be improved. As multipath period is longer for a static user than for a moving one, averaging time required to remove multipath is longer in the static case. This is demonstrated through simulated and live GNSS signals.