Fuzzy Receiver Design for GSM application

Abstract

This paper presents the application of nonlinear channel equalizers, based on fuzzy IF-THEN rules, for Global System for Mobile (GSM) communication applications. Two types of fuzzy filters namely Type–1 TSK (Takagi – Sugeno – Kang) fuzzy adaptive filter (FAF) and Type–2 TSK FAF are considered for performance comparison. The equalizers proposed can be trained with very short training signal of 26samples as desired by GSM data format. These equalizer performances are also compared with Bayesian equalizers based on maximum a-posteriori criteria and radial basis function (RBF) network Bit error rate (BER) is considered as performance index. The equalizer proposed here outperforms linear equalizers and other forms fuzzy equalizers in terms of BER at given SNR.