A novel statistical approach to ISI mitigation

Any communications signal suffers some degree of distortion during transmission. In high speed data communications, a significant portion of the distortion is ascribed to the presence of inter-symbol interference (ISI). Generally, an ISI channel can be modeled as a FIR filter. The ISI effect can thus be removed by inverse filtering the distorted signal at the receiver. The method presented in this paper, however, adopts a different equalization principle, which, rather than attempting to recover the original signal directly, adapts the decision rule in accordance with the channel statistics to achieve the same objective of retrieving the loss of BER performance.