A Robust Algorithm for Multichannel Eeg Compressed Sensing with Mixed Noise

Compressed Sensing (CS) has been widely used for telemonitoring of multichannel electroencephalogram (EEG) signals through wireless boday-area networks. However, most of existing multichannel EEG CS algorithms have not taken the noise into consideation or only considered the Gaussian noise. In this paper, we propose a robust multichannel EEG CS algorithm based on sparse and low rank representation in the presence of mixed noise (SLRMN). Our proposed algorithm involves an optimization model that takes both the Gaussian noise and the implusive noise into consideration, and the alternative direction method of multipliers (ADMM) is also developed to solve the proposed SLRMN. Moreover, we apply our method to EEG database to demonstrate the dramatic improvements in signal recovery compared to the state-of-the-art multichannel EEG CS methods, especially in the presence of mixed noise.