Multipath Interference Order Selection and Mitigation in GPS-Based Attitude Determination Systems

Abstract

INTRODUCTIONAttitude determination systems are important in many fields, such as land static measurements (geodesy) and mobile applications, as well as in satellites for communications. While attitude finding was traditionally based on inertial navigation systems (INS), in the last decades several navigation equipment manufacturers have developed GPS-based solutions for the problem of orientation finding. GPS-based attitude determination systems have been tested for land, marine, air, and space vehicles [1 -4]. These systems have succeeded in providing comparatively low-cost and highly stable solutions.Multipath effects on GPS receivers for positioning applications, such as geodesy, are investigated and reported in [5]. A geometrical model is presented and applied to reduce the multipath interference in a static receiver utilizing multiple antennas [6].It is well known that the accuracy of a GPS-based attitude determination system is considerably limited by multipath interference [7 -10]. A GPS-based attitude determination system is basically a differential phase measurement apparatus. Multipath affects the receiver operation when a few delayed echoes of the desired signal are received and combined with the desired signal. The signal mixture induces phase error, which results in turn in large navigation errors.A method for calibrating multipath interference is presented in [8]. The method of [8] is based on an approximation of the multipath interference using a sum of spherical harmonic functions. A different mitigation technique, based on the assumption of low dynamics, is presented in [9]. In [9], the multipath interference effect on the attitude measurements is modeled by a sum of sinusoids.A different approach to the problem is presented in [10], where the multipath plus thermal noise is modeled as a first-order Gauss-Markov process. This statistical model is suitable for applying a Kalman filter [11]. Both methods [9,10] assume that the multipath model order is known a priori.Recently, a new method for calibrating multipath interference for a GPS-based attitude determination system was proposed [12]. This method, known as the Sky Map Method, shows improvement in and is especially suitable for a satellite in orbit. This is due to a relatively static environment surrounding the receiver antenna array. In this case, the platform often experience a periodic dynamics, which makes the problem setting similar to a static scenario.A new postprocessing iterative algorithm for multipath interference mitigation is presented in this paper. The proposed algorithm is based on separation of the measured attitude into time sections where the multipath interference acts uniformly.