Helicopter radar return analysis: Estimation and blade number selection

The problem of helicopter parameter estimation and blade number selection is addressed. The parameters are estimated based on Maximum Likelihood methods, and the corresponding Cramér-Rao bounds are derived. To avoid ambiguities in blade number estimation, we apply information theoretic criteria for blade number selection incorporating alternate penalty functions. We assume that a clutter filter essentially removes all of the clutter and also part of the signal returns, lowering the effective signalto-noise-ratio. The Fractional Fourier Transform is used to separate the combined tail and main rotor signals into two different returns, allowing initial estimates of their corresponding rotor parameters. The proposed technique is validated by using returns from a helicopter observed experimentally with a pulseDoppler radar. KeywordsMicro-Doppler, Rotor blades, pseudo maximum likelihood, model selection, time-frequency, radar.