On Limits of Wireless Communications in a Fading Environment when UsingMultiple Antennas
Wireless Personal Communications: An International Journal
Joint transceiver design for MIMO communications using geometric mean decomposition
IEEE Transactions on Signal Processing - Part I
IEEE Transactions on Signal Processing
The intersymbol interference channel: lower bounds on capacity and channel precoding loss
IEEE Transactions on Information Theory
Achievable rates for Tomlinson-Harashima precoding
IEEE Transactions on Information Theory
Optimum power control over fading channels
IEEE Transactions on Information Theory
IEEE Journal on Selected Areas in Communications
A new THP precoding scheme with effective channel optimization
GLOBECOM'09 Proceedings of the 28th IEEE conference on Global telecommunications
Hi-index | 0.01 |
Tomlinson-Harashima Precoding (THP) can be combined with geometric mean decomposition (GMD) to decouple a multiple-input multiple-output (MIMO) channel into multiple single-input single-output (SISO) subchannels with identical signal-to-noise ratios (SNRs). The combined system is called GMD-THP MIMO system. As all subchannels for this system have identical SNRs, it is more convenient to design modulation/demodulation and coding/decoding schemes than other MIMO systems that have different SNRs among the subchannels. In this paper, we consider low-density parity-check (LDPC) coded GMD-THP MIMO systems. Modulo operation at the receiver is needed for THP decoding but it may generate modulo errors. These modulo errors cause the log likelihood ratio (LLR) values provided by conventional soft demapper to be very inaccurate. We propose an expanded soft demapper scheme to reduce the inaccuracy of the LLR values, by which significant performance improvement can be achieved for LDPC decoding. Furthermore, THP introduces power loss and modulo loss into the system. These two losses are related to the size of the modulo boundaries and therefore we derive the expressions for these two losses as functions of the modulo size factor. With these expressions, we find the optimal modulo size factor which achieves the minimum combined losses through doing a tradeoff between power loss and modulo loss. Computer simulations are presented to show that the tradeoff indeed improves the performance of the LDPC coded GMD-THP systems.