Joint DOA and TOA Estimation for Multipath OFDM Signals Based on Gram Matrix

Among the existing methods for joint estimation of direction of arrival (DOA) and time of arrival (TOA) in OFDM system, such as multiple signal classification (MUSIC) and estimating signal parameters viarotational invariance techniques (ESPRIT), sufficient sampling snapshots are required to achieve high precision. In this paper, we first analyze the constituent parts of the gram matrix of received signals under multicarrier scenario. While estimating DOA, it is proved that received signals from different subcarriers can be approximately regarded as multiple snapshots, received signals from different antennas are utilized for TOA estimation likewise. However, a certain amount of disturbance is introduced due to the cross correlation components. Thus, an improved eigenvector (EV) method is proposed to mitigate such adverse impacts. Simulation results are provided to demonstrate the effectiveness of the proposed method.