Semi-blind channel estimation and data detection for OFDM systems over frequency-selective fading channels

This paper considers semi-blind channel estimation and data detection for OFDM systems over frequency-selective fading channels. Using the maximum likelihood (ML) principle, we derive a blind channel estimator by taking the time domain transmitted signal as Gaussian (due to the central limit theorem) and averaging the likelihood function over the resulting Gaussian distribution. This estimator is realized using the steepest descent algorithm. Similarly, our semi-blind data detector integrates the channel impulse response (CIR) out of the likelihood function, which is realized using sphere decoding and V-BLAST. Simulation results show that our proposed channel estimator and data detector perform a fraction of dB within an ideal reference receiver.