Semi Blind Bussgang Equalization for Sparse Channels

This paper addresses the problem of semiblind equalization for long sparse channels using a sparse equalizer. We formulate the equalization problem following a Bussgang approach, derive the coherent MMSE input synbol estimator, and assess the performances of the analyzed semi-blind Bussgang equalizer in typical sparse channels. The simulations show that the semi-blind linear equalization algorithm allows flexible integration of the training in the eqaulization procedure while the sparse, not connected, support favourably copes with channels characterized by long delayed echoes.