Passband Equalization of Differentially Phase-Modulated Data Signals

Recently much attention has been given to high-speed synchronous data transmission via differential phase modulation and comparison detection. High-speed (above 2400 bits per second) transmission usually requires equalization 1 to compensate for the linear distortion introduced by the channel. This study will be concerned with the automatic and adaptive equalization of such data signals. At first glance, the nonlinear nature of the modulated signal would seem to preclude linear compensation. Upon closer examination it becomes apparent that a digitally phase-modulated signal, when resolved into in-phase and quadrature components, is linear in each component. Since this property is preserved after transmission through a linear medium, simultaneous linear equalization of each signal component becomes feasible. Due, however, to the purely quadratic nature of the channel219 220 T H E B E L L SYSTEM T E C H N I C A L J O U R N A L , F E B R U A R Y 1 9 7 3 dependent terms present in the differential detector output, a linear equalizer should 1 precede the detector-hence the equalizer operates in the frequency passband. Resolving the channel output into in-phase and quadrature components suggests that satisfactory performance will be obtained if the equalized in-phase signal is approximately Nyquist and the equalized quadrature signal has most of its samples close to zero. This is accomplished by choosing an appropriate cost function (of the equalized samples) and using an appropriately structured equalizer capable of minimizing this function.