A cumulant interference subspace cancellation method for blind SISO channel estimation

We present a new linear method for blind estimation of single-input single-output (SISO) finite impulse response (FIR) channel. The channel can be minimum-phase or nonminimum-phase channel. The proposed method is based on a series of fourth order cumulant matrices, where it is shown that by employing vectors chosen from the left space of a concatenated cumulant matrix, the interference subspace of the channel convolution matrix can be cancelled and thus, channel information can be extracted. The proposed method is robust to channel order overestimation, and it has a similar computational complexity as other existing methods. Simulation results are included to validate the performance of the proposed algorithm.