Direct position determination of OFDM signals

Direct Position Determination (DPD) has recently been proposed [1] as a high-accuracy technique for emitter localization. It has been shown that DPD outperforms traditional emitter localization methods such as Angle-Of-Arrival (AOA) and Time-Of-Arrival (TOA) as it solves the localization problem using the data collected at all sensors at all base stations together, as opposed to the traditional AOA/TOA approach that is composed of two separate steps: 1) AOA/TOA independent estimates and 2) triangulation based on the results of the first step (see e.g., [2]). The signal model in [1] assumed a narrow-band Gaussian source with a rather simple flat-fading channel model. Here we extend the DPD algorithm to OFDM signals that propagate through a frequency-selective channel and derive the Maximum Likelihood Estimator (MLE) for the problem.