A Spectrally Efficient Acquisition Algorithm for MIMO-OFDM

Abstract

This paper proposes a spectrally efficient algorithm for acquisition of symbol timing, carrier frequency and channel estimation in MIMO-OFDM. Depending on the MIMO constellation size, the proposed method attains those tasks with a preamble that is 8 to 16 times shorter than the one specified by the IEEE 802.11a/n standards. The algorithm can acquire frequency offsets that span the entire transmission band. Simulations show that the outage performance due to frequency acquisition errors is below 0.1% for all cases of interest. It is also shown that the algorithm attains ideal symbol timing in over 96% of the cases with two or more transmit branches in SNR above 7 dB. The remaining fraction of transmissions end up with degraded channel estimates, although the required error rate performance may still be attained.