On the application of the LMS-based adaptive noise canceller in nonstationary environment associated with airborne Doppler weather radar

The use of the LMS least mean square-based adaptive noise canceller (ANC) as a process decorrelator in a nonstationary signal and noise environment is investigated. In an airborne Doppler weather radar application, ground clutter and weather returns are treated as two uncorrelated nonstationary processes. Under severly low signal-to-clutter ratio conditions, the adaptive noise canceller exhibits performance superior to that of a fixed, nonadaptive system in terms of ground clutter suppression and estimates of weather parameters. Signal modeling and the choice of reference input, which are issues associated with the implementation of the ANC, are also addressed.<>