EQUALIZATION - Advanced Signal Integrity for High-Speed Digital Design - page 542

precursor and three postcursor taps. Calculate the sample values of the
resulting equalized response.

− 75

− 74

− 74

− 68

18

− 41

− 62

y = − 68 mV

− 71

− 72

−

72

− 73

− 73

− 74

− 74

12-4 Using the received pulse response from Problem 12-3 and the coefficients
for the DLE and DFE below, calculate the sample values of the equalized
output response.

0 . 35

0 . 010

0 . 005

0 . 65

= =

C DLE C DFE

−

12-5 Derive the transfer function for a discrete linear equalizer, starting from
equation (12-21). The time-shift property of the Fourier transform will
prove useful: If y ( t ) has transform F [ y(t) ] (f ) = Y(f) , then
∞

y(t − t 0 )e − 2 πjf t dt = e − 2 πjf t 0 Y(f)

−∞

12-6 Derive the transfer function of the alternate passive equalizer shown in
Figure 12-20.
12-7 An alternative to the ZFS approach is the minimum mean square
error (MMSE) equalizer. Finding the coefficients using the MMSE
requires first calculating the autocorrelation matrix, R xx x T x , and the =

x T y . The equalizer coefficients are then cross-correlation matrix, R xz =

R − 1

xx R xz . Use the pulse response input samples from
the Problem 12-3 to calculate the coefficients for a linear equalizer with
one precursor and three postcursor taps using the MMSE algorithm.
Calculate the sample values of the resulting equalized response.
12-8 Given the pulse response in Problem 12-3, calculate the equalizer coef-
ficient progression obtained using the adaptive ZFS algorithm.

calculated as c =