Low-complexity fuzzy filter-assisted turbo equalization for dispersive Rayleigh-fading channels

Turbo equalisation (TEQ) is an iterative detection and decoding technique that has been shown to be successful in mitigating the effects of intersymbol interference (ISI) introduced by partial response modems and by dispersive channels. In this paper, we propose a fuzzy filter-assisted decision-feedback equaliser (DFE)-based TEQ scheme that not only is comparable in performance to the Jacobian radial basis function (RBF) DFE-based TEQ but also is low-complexity. The per iteration computation of both the fuzzy filter DFE TEQ using product inference and maximum defuzzification and the fuzzy filter DFE TEQ using minimum inference and maximum defuzzification is much lower than that of the Jacobian RBF DFE TEQ for binary phase-shift keying (BPSK) modulation, 4 quadrature amplitude modulation (QAM), and 16 QAM. Notably, the computational complexity of the Jacobian RBF DFE TEQ is exponentially related to the feedforward order, but the fuzzy filter DFE TEQ using minimum inference and maximum defuzzification is independent of that. Previous TEQ research has shown that the Jacobian RBF DFE TEQ considerably reduces the computational complexity compared to the logarithmic maximum a posteriori (Log-MAP) TEQ with similar performance. In this present study, the proposed fuzzy filter DFE TEQ further reduces the computational complexity with slight performance degradation in contrast to the Jacobian RBF DFE TEQ. With this, the materialisation of the TEQ scheme becomes more feasible. Copyright © 2007 John Wiley & Sons, Ltd.