Iterative subspace alternating projection method for GNSS multipath DOA estimation

Abstract

The signal direction of arrival (DOA) estimation is an important research topic in GNSS multipath mitigation. Correct estimation of the number of multipath echoes as well as their DOAs is a critical information for the following mitigation techniques. In GNSS multipath scenarios, alternating projection method has been proved efficient for estimating coherent incident signals. It has shown asymptotically consistent performance as maximum likelihood estimator but computationally improved from multi-dimensional searches to one dimension. The AP method initially assumed the number of sources known, and the source number detection task was accomplished by other methods such as minimum description length principle. In this study, based on the AP method, a set of iterative updating formulas are derived to replace the one-dimensional search, so further decreases the computational complexity. A source number detection strategy for coherent signal is also proposed. This strategy is designed compatible with the tracking loop architecture of GNSS receiver, so bringing simplified implementation and accuracy improvements at low signal to noise. Simulation results illustrate the superiority of the proposed method.