IEEE 802.11ac WLANs: A performance evaluation of sphere decoding and lattice reduction MMSE-SIC MIMO detectors

Abstract

The 802.11ac amendment, approved by the IEEE in late 2013, is an evolution of the 802.11n amendment and specifies a physical (PHY) layer based on orthogonal frequency division multiplexing (OFDM) multiple-input multiple-output (MIMO) with up to eight spatial streams (SS). The challenges and issues regarding the analyzes, design and optimization of sphere decoding (SD) and lattice-reduction mean squared error successive interference cancellation (LR MMSE-SIC) MIMO detectors in 802.11ac wireless local area networks (WLANs) are investigated in this paper. This industrial oriented paper shows simulation results that take into account fundamental operational aspects for baseband design of 802.11ac PHY layer (e.g., synchronization, channel estimation, coding) operating over spatial and frequency correlated Task Group ac (TGac) channel models. Binary convolutional code (BCC) with hard-decision Viterbi decoding is assumed in this paper. For both single-user (SU) and multi-user MIMO (MU-MEVIO) environments, we show that the SD MIMO detection allows power gains between 1 and 3 dB in relation to LR-based MMSE-SIC MIMO detection. Hence, since the SD algorithm presents a computational complexity that depends on the signal-to-noise ratio (SNR), while the LR-based MMSE-SIC MIMO detector has a fixed computation burden for a fixed number of SS and receive antennas, we have concluded that an IEEE 802.11ac PHY layer based on software-radio (SR) must have an optimized scheduling core to set on the fly the best MIMO detection scheme based on complex tradeoffs between performance and computational complexity.