Acoustic echo cancellers are integrated into various speech communication devices, such as hands-free conferencing systems and speakerphones. Microphone arrays can be employed to enhance the performance of such systems, though they assume a static environment when transitioning to double-talk, and rely on double-talk detection. This work introduces a multichannel echo canceller implemented by a microphone array beamformer that can adapt to a changing environment where the locations of both the farend and near-end sources change during double-talk, with no double-talk detector. This is done by utilizing multiple recent frames in the short-time Fourier transform (STFT) domain. We show how can the acoustic paths be accurately estimated given the recent time frames of the far-end and microphone signals. Also, our beamformer aims to reduce background noise. Simulations are conducted in a reverberant room with nonlinear loudspeaker distortion and realistic low signal-to-echo ratio (SER) resembling a speakerphone. The experiments demonstrate the advantages of the proposed approach compared to normalized leastmean-squares (NLMS) based approaches.INDEX TERMS Acoustic echo cancellation, array signal processing, beamforming, adaptive filtering.