Interpolation-based Matrix Arithmetics for Mimo-ofdm Systems

Multiple-input multiple-output (MIMO) techniques, characterized by the use of multiple antennas at both the transmitter and the receiver, are a key element in the design of future wireless communication systems. With respect to single-antenna systems, MIMO systems offer higher spectral e ciency through spatial multiplexing gain, improved link reliability through spatial diversity gain, and greater coverage through array gain. For transmission over frequency-selective channels, MIMO techniques are typically combined with orthogonal frequency-division multiplexing (OFDM), which drastically simplies channel equalization by transforming a wideband MIMO channel into a set of parallel narrowband MIMO channels (corresponding to OFDM tones), over which data transmission can be carried out independently. Detection algorithms for MIMO-OFDM wireless systems typically require computationally expensive matrix operations (such as matrix inversion, QR decomposition, or LU decomposition) to be carried out on the MIMO-OFDM channel matrices corresponding to the datacarrying tones. This procedure, referred to as preprocessing, exhibits in general a signi cant computational complexity, since in practical systems the number of data-carrying tones ranges from 48 (as in the IEEE 802.11a/g standards) to 1728 (as in the IEEE 802.16e standard). The straightforward approach to MIMO-OFDM preprocessing, consisting of brute-force per-tone application of the matrix operation, ignores the fact that the MIMO-OFDM channel matrices result from