A Comparison of Receivers Designed on the Basis of Minimum Mean-Square Error and Probability of Error for Channels with Intersymbol Interference and Noise

In much theoretical work, the minimum mean-square error (MMSE) criterion is used as a basis of design for receivers of dispersed digital signals in the presence of noise. A variant of this criteria that is also frequently used is the MMSE with the constraint of zero intersymbol interference (OISI). In this discussion we compare the probability of error
P(e)
of receivers designed with the MMSE and MMSE-OISI criterion with a receiver designed by computer minimization of
P(e)
by variation of pole locations of a fixed order filter. The performance of these receivers is then compared with the isolated pulse lower bound and the probability of error of the maximum likelihood receiver. These Comparisons are made for a channel that is an approximation to 6000 ft of twisted pair paper insulated cable with a transmission rate of 1.5 Mb.