The detection performance of multi-level decision feedback equalizatio
n (MDFE) and partial response with maximum likelihood detection (PRML)
in the presence of magnetoresistive (MR) head nonlinearity is investi
gated through simulation. Two models are used to characterize the MR h
ead nonlinearity: (1) a transfer function curve based on measurement a
nd (2) a 3rd-order polynomial model. The results indicate that MDFE is
more robust to MR head nonlinearity than PRML, especially to head sat
uration.