MMSE Turbo Equalization and Detection for Multicarrier Faster-Than-Nyquist Signaling

The Evolution of Faster-Than-Nyquist Signaling

“…Peng et al [102] Developed a clipping-based PAPR reduction scheme for MFTN signaling, while the symbol packing ratio, the subcarrier spacing, and the clipping ratio maximizing the bandwidth efficiency are numerically searched. Peng et al [104] Proposed an MMSE-based equalization scheme combined with SIC for MFTN signaling, which achieves the BER performance close to that achieved by the maximum a posteriori probability equalization scheme combined with SIC with lower computational complexity. Tian et al [105] Proposed a factor-graph-based receiver for MFTN signaling, where a Gaussian message-passing (GMP) algorithm was developed for the derived factor graph representation of MFTN signaling.…”

“…Note that MFTN signaling was also referred to as time-frequency packing [91,96]. The TD transmit MFTN signal is represented by [97,104,105]…”

“…In [91], Barbieri et al derived the constrained information rate of MFTN signaling employing a realistic discrete modulation scheme, showing that when using an RRC shaping filter and a Gaussian shaping filter, MFTN signaling achieves a higher bandwidth efficiency than its orthogonal counterpart. An efficient detection scheme for MFTN signaling has been developed in [90,91,93,97,104,105,109], which considers both the effects of ISI and ICI. Moreover, the peak-to-average power ratio (PAPR) problem of MFTN signaling caused by multi-carrier transmission has been considered in [100-102, 108, 110].…”

Parametric Bilinear Iterative Generalized Approximate Message Passing Reception of FTN Multi-Carrier Signaling

“…In [5], a maximum a posteriori (MAP) probability receiver based on the Bahl-Cocke-Jelinek-Raviv (BCJR) algorithm [10] was proposed for approaching the bit error rate (BER) performance of Nyquist systems, albeit at the expense of an exponentially escalating complexity. An attraction 2D minimum meansquared error (MMSE)-based equalizer was designed for the interference mitigation of MFTN signaling in [11], where authors succeed in rendering the complexity independent of both the number of subcarriers and of the constellation size. However, its BER performance was significantly degraded in severe ICI scenarios.…”

Channel Estimation Based on the Discrete Cosine Transform Type-III Even

“…In any case, MCM systems require the estimation of the channel's impulse response (CIR), which is usually unknown. This estimation is typically achieved by using a known training symbol, and several authors have addressed channel estimation and equalization in OFDM [7]- [10]. Similarly, other works have provided estimation algorithms for different types of DCTs.…”