Interference Localisation Trials Using Adaptive Antenna Arrays

This paper presents the results of an interference localisation trial using an experimental adaptive antenna array currently under development. The prototype adaptive antenna array was driven through a field of RF transmitters and the direction of arrival (DOA) of multiple RF signals at various positions of the vehicle were estimated. From these DOA measurements the transmitter locations were estimated by triangulation. The adaptive antenna array was also used to reject the interfering RF signal in real time to allow the accurate GPS position of the vehicle to be determined in the presence of strong interferences. INTRODUCTION A number of methods have been proposed to localise GPS interferences. The drop in C/No values on the GPS receiver can give an estimate of the interference signal strength and by measuring the C/No values on the GPS receiver at various locations or platform attitudes, a rough estimate of the interference location can be obtained [Geyer 1999]. More sophisticated approaches capture the raw data from several GPS antennas. There are two main methods, depending on the separation of these antennas. If the antennas are separated by long distances, the time difference of arrival (TDOA) of the interference signal at each antenna can be used to locate the interference source directly by triangulation [Gromov, 2000]. For closely spaced antennas, as is the case in an adaptive antenna array, the phase information on multiple antennas can be used to give an estimate of the interference Directions of Arrival (DOAs), estimated at various locations of a moving platform these DOAs can then be used to determine the interference source location. This approach was used in these trials. The dual use of GPS adaptive antenna arrays for both interference mitigation and localisation has been studied in previous papers. In [Falcone, 2001], the statistical fluctuations of the DOA estimation errors are considered as a function of interference strength and signal averaging time. These results give an indication of the required signal averaging times to obtain accurate DOA estimates of low power interference signals. These results are based on simulation and laboratory tests and do not appear to include all the non-ideal effects in the actual antenna array. Some further experimental DOA estimation results from a three-element experimental patch antenna array are given in [Bond, 2000], these results were based on open air radiation tests and thus included the non-ideal effects of the antenna array. In [Trinkle, 2001a] DOA estimation results from an eight element monopole antenna array are presented, along with experimental results from localising a single interference source. In this paper a scenario in which multiple interference sources are to be localised is considered. The paper is structured as follows: Firstly, the system hardware is described including the actual antenna array and data logging facilities. Next, the issue of array calibration is briefly considered. The observed phase errors in the array are discussed and a simple array calibration procedure is applied to compensate for these errors. The expected accuracy of the interference DOA estimates obtained from the calibrated array are then estimated using the Conventional Beamformer (CBF), the Minimum Variance Distortionless Response Beamformer (MVDR) and MUSIC (see [Therrien, 1992]). Finally the results of an interference localisation trial are discussed, in which the antenna array was driven through a field of RF transmitters.