Comparison between LS and TLS in adaptive processing for radar systems

Electromagnetic noise like interference (NLI) disturbs the proper working of radar by injecting noise from specific angular directions. Phased-array radars contrast this phenomenon via adaptive spatial filtering of the interference; this objective is reached by distorting the receiving antenna pattern by generating a null in the interference direction of arrival (IDoA). In the real world the adaptive cancellation of directional interference is limited, among other error sources, by the presence of clutter and by the amplitude and phase mismatch of receiving channels; thus adaptive signal processing strategies have to be conceived to permit the NLI cancellation in real environment. This paper compares two methods for adaptive signal processing in modem radar systems. We recall the working principles of the least squares (LS) and total least squares (TLS) algorithms. These processing methods are applied to three study cases in adaptive phased-array radar. We study the estimation of the jamming covariance matrix in presence of clutter in an adaptive phased-array. The null depth in adaptive jammer cancellation is then analyzed, and the problem of equalization of receiving channels is considered.