Iterative (turbo) estimation and detection techniques for frequency-selective channels with multiple frequency offsets

We propose an iterative channel estimation and data detection technique for frequency selective channels with multiple frequency offsets (FOs). The pilot symbols are generally inadequate to obtain an accurate estimate of the FOs due to limitation on the frequency resolution of the estimator. Therefore, we initially use the pilot sequence for the estimation and equalization of the channel without consideration to FOs. However, we then use the soft estimate of the transmitted signal as a long pilot sequence to determine multiple FOS and to refine channel estimates iteratively. The proposed iterative technique also has the ability to resolve multipaths, bringing the multiple FO problem into estimation of distinct harmonics. We verify the detection performance using the bit-error-rate curves and the estimation performance through comparison of the variance of estimates with the Cramer-Rao lower bounds.