Nonlinear equalization for super-resolution near-field structure discs

Abstract

In this paper, we present a neural network-based equalizer (NNEQ) for super-resolution near-field structure (super-RENS) discs. In order to investigate the presence of nonlinear interactions, we employ the bicoherence test, which is based on higher-order statistics. This test reveals that a super-RENS disc experiences severe nonlinear inter symbol interference (ISI). To mitigate the nonlinear ISI, we apply the NNEQ, which relies on the nonlinear autoregressive network with an exogenous input (NARX) model. Its validity is tested with radio frequency (RF) signal samples obtained from a super-REVS disc. The performance of the proposed equalizer is superior to that of the case without equalization and that of the case with limit-EQ in terms of bit error rate.