Abstract: A high resolution blind frequency estimation method for multi-component modulated signal in distributed multiple input multiple output was proposed in this paper. Did not need the prior information of pilots and avoiding sparse represent of signals, the multiple frequency offsets could be estimated once without the interpolation and phase unambiguity in traditional frequency estimation methods. The estimation of the transmitted signals from the received signals could be obtained by using the determined blind source separation according to the characteristic of distributed MIMO systems. After demodulation processing, the signals were transformed to multi-component single frequency signal. According to the sparse representation of the signal, the proposed method utilized a random compressed matrix and estimated the signal frequencies in the compressed domain by -norm optimization. Simulation results show that, the proposed method can achieve high resolution frequency estimation with large reduction in computation at low SNR comparing with existing methods. The average mean square error can achieve 1e-6 at 5dB.