DFT-Based Channel Estimation and Noise Variance Estimation Techniques for Single-Carrier FDMA

Practical frequency domain equalization (FDE) systems generally require knowledge of the channel and the noise variance to equalize the received signal in a frequency-selective fading channel. Accurate channel estimate and noise variance estimate are thus desirable to improve receiver performance. In this paper we investigate the performance of the denoise channel estimator and the approximate linear minimum mean square error (A-LMMSE) channel estimator with channel power delay profile (PDP) mismatch and noise variance estimation errors for a single carrier frequency division multiple access (SC-FDMA) system. A windowed DFT-based noise variance estimator is also presented in this paper to improve A-LMMSE channel estimation as well as to facilitate the MMSE-FDE design. Results show that the proposed noise variance estimator remains unbiased up to a SNR of 50dB.