Non uniformly spaced equalization

Depending on the sampling rate, two types of equalizers have been extensively studied in the past: i) baud spaced equalizers (BSE) and ii) fractionally spaced equalizers (FSE). In each case, uniformly spaced samples of the received signal are processed by the equalizer. In this paper, we present a new approach to equalization where the equalizer operates on periodic nonuniformly spaced samples of the received signal. We first introduce a continuous-time ISI channel model with a periodic nonuniform sampler and then, using appropriate multirate DSP building blocks, we derive an equivalent multichannel discrete-time model. Necessary and sufficient conditions for perfect channel equalization using FIR filters are obtained. These results indicate that for a large class of communication channels, the newly proposed equalization scheme enjoys the combined benefits of BSE and FSE.