Channel equalization represents a defensive strategy against temporally dispersive channels. It created a splendid history of data transmission by enabling the data rate on a twisted line telephone to increase from 1.2 to 19.2 kb/s shortly in two decades from 1970s to 1980s. When a pulse sequence is transmitted over a dispersive channel, the pulse waveforms are spread over time, producing an overlap between the overall channel impulse response and its translates. All the dispersive symbols compete for use of the same channel, leading to a phenomenon referred to as inter-symbol interference (ISI). The influence of ISI can be minimized based on different criteria, resulting in different types of channel equalizers, which are studied in this chapter. A beautiful picture exists that links channel equalization as an engineering practice to the power spectral decomposition theory in stochastic processes.
For digital communication in a high-data-rate environment, the propagation is usually modeled as a frequency-selective channel. The channel behaves like a filter, which can be described by a channel matrix in the discrete-time domain or, equivalently, by a nonideal bandlimited frequency transfer function in the frequency domain. This ...