Detecting aircraft in auroral clutter by HF sky wave radar using 2D continuous wavelets

Operation of sky wave radar in Canada is challenged by the presence of ionospheric clutter from the earth's auroral zone due to severe Doppler spreading, which results in clutter-limited detection of aircraft. Elevation angle control in the transmit and/or receive beams to separate the target can mitigate the auroral clutter to some degree, but may not be always effective especially when operating at high latitudes or near polar regions. This paper reports using the two-dimensional (2D) continuous wavelet transform (CWT) analysis to `unearth' a target buried in the auroral clutter. The approach is based on the ability of the CWT to provide detailed local spectral information about a signal for arbitrary scale, orientation, and physical location. Results obtained using real clutter signals with simulated targets are reported. Very promising initial results, which show an improvement of the signal-to-clutter ratio (SCR) of 45-58 dB, are presented. However, a closer look indicates that these `target detections' are in reality due to wavelets' ability to detect extremely small signal discontinuities introduced during the embedding of synthetic targets in the range compressed data. When inserting a synthetic target directly in the raw data before the range compression, the SCR improvement is shown to be only marginal.