Sub-Block Noncoherent Space-Frequency Coding with Full-Diversity for MIMO-OFDM

Abstract

Space-frequency coding is an attractive approach to exploit the space and frequency diversity provided by an orthogonal frequency-division multiplexing (OFDM)-based frequency-selective multiple-input multiple-output (MIMO) fading channels. Due to the prohibitive complexity of acquiring knowledge of the fading coefficients, noncoherent space- frequency coding (NSFC) which dose not need the knowledge of channel is proposed and the design criteria is presented. However, the existing NSFCs adopt the scheme that one block occupies all the subcarriers of the OFDM system, and this large code size increases the coding and decoding complexity exponentially. In this paper, we address a new view of the transmission of space-frequency codes which transforms the transmission in frequency selective channels into flat fading channels. Then, a sub-block noncoherent space-frequency coding is proposed which divides the space-frequency plane into small sub-blocks and each one constitutes a space-frequency codeword. This reduces the code size and the coding-decoding complexity. By a modified signal model, an asymptotic analysis of the pairwise error probability (PEP) is derived and thus the design criteria for full diversity-achieving code is explicitly defined. We also propose a code construct that achieves the promised order of diversity and demonstrate our conclusion by computer simulation.