Optimal Training for the SM-MIMO-OFDM Systems with MMSE Channel Estimation

Spatially multiplexed multiple-input-multiple-output orthogonal frequency division multiplexing (SM-MIMO-OFDM) systems are expected to provide a considerable increase of transmission capacity paving the way for the next generation broadband wireless communication standards. Considered MIMO-OFDM receiver performs estimation of the channel state information based on transmission of pilots concurrently with data symbols. The channel estimates are subsequently used for data detection. In this paper we discuss optimal placement and number of pilot subcarriers and derive the closed-form expression for the optimal power allocation between pilot and data symbols. The latter factor is important as under total transmit energy constraint there is a trade-off between boosting power of the pilot symbols to gain more accurate channel estimates and maintaining higher signal-to-noise ratio at the data subcarriers to minimise probability of detection error in the assumption of sufficiently reliable channel state information.