High data-rate transmission with eigenbeam-space division multiplexing (E-SDM) in a MIMO channel

Abstract

When channel state information is known at a transmitter in multiple-input multiple-output systems, the optimum capacity is given by eigenmode channel division with water-pouring power control. In this eigenbeam-space division multiplexing (E-SDM), bit assignments to substreams based on the capacity is not optimum due to the fact that the number of assigned bits is expressed by discrete quantity. In the paper, a method to assign both bit and transmit power to each substream based on the criterion minimizing total bit error rate (BER) is developed, and the BER performance is numerically analyzed in comparison to spatial division multiplexing (SDM). The simulation results assuming 5-transmit and 2-receive antennas show that the E-SDM provides about 10 dB gain compared to the conventional SDM at average BER of 10 −3 . 0-7803-7467-3/02/$17.00 ©2002 IEEE.

Instantaneous Signal and Self-Interference Power of MIMO Eigenmode Transmission with Feedback Time Delay

“…However, as discussed earlier, regardless of the feedback delay, the SINR degrades whenever two or more eigenvalues are nearly the same. The results for (2,2), (2,4) and (4,4) systems are shown in Figs. 4, 5 and 6 respectively.…”

“…This has inspired the use of singular value decomposition (SVD)-based MIMO transceiver architectures [2] [3]. Since multiple independent data streams can be transmitted simultaneously through these uncoupled eigenmodes, the overall data rate can be improved significantly [4]. Nevertheless, this kind of system requires the transmitter to obtain channel state information (CSI) from the receiver with the aid of a feedback link.…”

Pseudo Eigenbeam-Space Division Multiplexing (PE-SDM) in Frequency-Selective MIMO Channels

“…The complexity in the PEV scheme was examined for the EVD of R rx (0) and the GS operation of H H F ( f )V 0,K , and that in the TEV scheme was examined for the EVD of H H F ( f )H F ( f ) for comparison, where EVD was solved by Jacobi method. Therefore, it should be noted that they do not contain calculation of FFT, matrix multiplication, and the summation in (13). For the TEV scheme, to speed up the convergence in EVD, U e ( f − 1) obtained at the adjacent frequency point f − 1 was utilized as an initial unitary matrix for U e ( f ).…”

“…We prepared four modulation schemes: QPSK, 16 QAM, 64 QAM, and 256 QAM. Under these conditions and total TX power constraint, the transmitter optimally determines modulation scheme and TX power per substream so as to give the minimum BER averaged over utilized substreams after Viterbi decoding [13], [34]. We employed mean eigenvalue E λ i ( f ) or mean pseudo eigenvalue E λ i ( f ) averaged over all subcarriers in each substream as the substream-quality measure in the decision.…”

Improvement of transmission performance by variable precoding in MIMO transmission

“…The signal separation methods are classified into two types: one where Channel State Information (CSI) is known at both the transmitter and receiver, and another where CSI is known only at the receiver. A representative example of the former is eigen-mode transmission method (E-SDM) [4], where the transmitted weights matching to CSI are multiplied by the transmitted signals. These weights are called precoding.…”