McKerracher, Moo, and Ponsford Figure 6. Example of traditional impulsive noise cancellation techniques. Impulsive noise observed between pulse 310 and 370. Corrupted data is either simply blanked by setting values to zero or values predicted. lope of the signal plus interference must be larger than typical in order to detect within the time segment. A typical example is shown in Figure 6. Radar systems in the HF band operate in an environment of high clutter and therefore have a large signal envelope in the time domain. Targets of interest may be present in the portion of the frequency (Doppler) domain that is well removed from the clutter, thus are noise floor limited. The effective signal to noise ratio of these targets may be adversely affected in this situation. Figure 7 and Figure 8 demonstrate an example where the time domain envelope is largely unaffected by the low level impulse event however the frequency domain noise floor is elevated by 20 dB, resulting in a significant loss in signal-to-noise ratio (SNR). The developed approach uses the wideband data as observed on the Spectrum Management system to identify pulses that are corrupted by impulse noise. The impulse event is significantly easier to detect in the wideband frequency domain rather than the time domain. The noise floor increase in the wide band frequency domain from a short duration impulse is shown in Figure 9. The wideband data collection and the narrowband target processing are time aligned. Thus, a corrupted sweep in the wideband data can be correlated to a specific pulse sequence in the narrowband target data channel. The identified pulse(s) in the narrowband channel can be replaced by zero padding or with interpolated data obtained from adjacent DECEMBER 2017 Figure 7. Pulse envelope (time domain) with low amplitude impulse event. Figure 8. Frequency domain of signal. Low amplitude impulse resulted in nearly 20 dB increase in noise floor. Figure 9. Wideband noise floor elevation due to impulsive noise events. IEEE A&E SYSTEMS MAGAZINE 33