On the performance evaluation of an LTE SFBC system with Wiener channel estimation in frequency-selective channels

The uncoded performance of an Alamouti space-frequency block code operating in a frequency-selective channel is studied in the presence of channel estimation errors. The Alamouti Maximum Likelihood (ML) symbol detector with linear combining is considered at the receiver. Imperfect channel estimates are provided by a realistic as well as robust Wiener interpolator. A closed-form expression of the uncoded error probabilities of a space-frequency block code employing an Alamouti with linear combining is derived. Our analysis is conducted in the framework of a Long-Term Evolution air interface. The impact of channel estimation errors on the uncoded performance of the system is evaluated as a function of key system parameters, such as the anticipated power delay profile at the channel estimator, the actual, i.e. true, power delay profile, the pilot spacing and the Wiener filter length. The role of partial knowledge of the channel statistics is investigated as well. Finally, we derive a lower bound from our closed-form expression and calculate the bit error outage of the linear Alamouti combiner in presence of log-normal fading signal-to-noise ratio (SNR).