A receiver signal‑processing operation that compensates or reverses intersymbol interference (ISI) and other linear distortion introduced by the transmission channel, producing an output sequence intended to match the transmitted symbol sequence sufficiently for reliable detection or decoding.

Principle

An equalizer implements an approximate inverse (linear or nonlinear) of the channel's effect over the communication bandwidth; its design balances ISI suppression against noise amplification, complexity and sensitivity to channel estimation errors.

Demonstration

Illustrative scenario → A baseband receiver facing multipath ISI applies a finite‑impulse‑response (FIR) linear equalizer whose coefficients were computed from a channel estimate. Recognition → Measured eye closure indicates ISI. Action → The equalizer filters the received samples to undo the channel's smearing. Consequence → The equalized waveform yields larger eye openings and lower symbol error rates when noise amplification is controlled.

Misapplication

Conflating equalization with channel estimation. Equalization is the processing that uses a channel estimate (or adaptive error signals) to remove ISI; treating them as the same step ignores that estimation provides parameters and equalization applies them.

Consequence

Effective equalization reduces residual ISI and improves symbol detection; however aggressive inversion can amplify noise or produce unstable filters if the channel is non‑minimum phase or the estimate is poor, so equalizer choice affects robustness, latency and implementation cost.

Reversal

When the channel has severe non‑minimum‑phase characteristics or when noise dominates the signal, simple inverse filtering increases error; in such cases sequence detection (e.g., Viterbi/MLSE) or decision‑directed nonlinear equalizers may outperform linear inversion.

Boundary

Clearly within: receiver modules designed to mitigate ISI from linear, time‑limited channel impulse responses (linear FIR equalizers, DFEs, MLSE). Boundary case: channels with mixed linear and nonlinear distortion where equalization must be combined with other compensation. Clearly outside: channel coding, source decoding, or RF front‑end filtering that are not intended to remove ISI on a symbol‑by‑symbol basis.

Semantic Tension

Suppression versus amplification: stronger ISI suppression (sharper inverse) reduces intersymbol interference but tends to amplify noise and increase sensitivity to estimation error; complexity versus latency: more sophisticated equalizers reduce residual errors at the cost of computation and delay.

Synthesis

Equalization is the operational inversion of channel distortion at the receiver: effective designs choose the inversion strength and architecture to trade residual ISI against noise amplification, computational cost and the available quality of channel state information.