Adaptive reduced-state sequence estimation for linearly and nonlinearly distorted signals in magnetic recording channels

An adaptive reduced-state sequence estimation (RSSE) is described, which uses a memory table to equalize both nonlinear distortion, a particular problem in high density recording channels, and linear distortion. Time-variant factors contributing to both types of distortion can be handled by updating the memory table. RSSE is far less complicated than MLSE, whose complexity results from long channel memory. Error probability analysis accomplished with computer simulation confirms the potential for excellent performance in high density magnetic recording systems.<
>