Constrained Adaptive Arrays for Narrow-Band GNSS Signal Receivers

Two types of adaptive antenna arrays for the Global Navigation Satellite Systems (GNSS) are considered in this paper. Both arrays use the constrained Recursive Least Square (RLS) adaptive filtering algorithms for the array weights calculation. The linear constraint is applied to keep a fixed value of a selected weight of the array or to keep the required radiation pattern value towards the information signal source. These constraints allow to avoid the usage of the desired signal, which is usually required in the adaptive filtering algorithms. The reason is that the GNSS information signals are masked by the thermal noise at the array output and this does not allow to estimate the error signals (the desired and array output signal difference) whose absolute values are minimized during the adaptation. The results of the arrays' simulation are presented. They demonstrate that both adaptive arrays can suppress the interference signals with approximately equal efficiency in the same conditions. However, only the beam constrained array is able to keep a required radiation pattern value in the direction of interest, if this direction is a priori known. The direction of interest is a navigation satellite angular location. However, since a GNSS terminal receives the signals from all the satellites in the line of sight, some of them are always received even in the case of the weights constrained array. These are the satellites which are located in the directions where the radiation pattern of the array does not degrade significantly during the adaptation.