The recent development of fast A/D converters and digital signal processors has considerably affected the modern radar system design. In the most popular configuration to date, the main channel (a conventional beamformer) is digitized along with a number of auxiliary channels. This configuration forms the basis for the adaptive sidelobe canceller (ASLC), which has been proposed for mitigating the influence of jammers that are present in the sidelobes of the main channel of the array. The ASLC can be efficiently implemented in real-time using recursive least-squares techniques, and has been demonstrated to perform well in certain scenarios. However, the ASLC has a number of shortcomings. The method fails, for instance when the target signal is too strong. This drawback can be eliminated by applying a parametric approach. Herein, the exact maximum likelihood (ML) estimator assuming a sinusoidal target signal is derived. The computational complexity of the ML estimator is found to be comparable to that of the ASLC. Initial simulation results indicate that the ML and ASLC methods perform similar at low SNR's, but that the ML estimator does not share the signal cancellation phenomena observed in thw ASLC.