Low-complexity iterative receiver design for high spectral efficiency communication systems

Publication Type:
Issue Date:
Full metadata record
With the rapid development of the modern society, people have an increasing demand of higher data rate. Due to the limited available bandwidth, how to improve the spectral efficiency becomes a key issue in the next generation wireless systems. Recent researches show that, compared to the conventional orthogonal communication systems, the non-orthogonal system can transmit more information with the same resources by introducing non-orthogonality. The non-orthogonal communication systems can be achieved by using faster-than-Nyquist (FTN) signaling to transmit more data symbols in the same time period. On the other hand, by designing appropriate codebook, the sparse code multiple access (SCMA) system can support more users while preserving the same resource elements. Utilisation of these new technologies leads to challenge in receiver design, which becomes severer in complex channel environments. This thesis studies the receiver design for high spectral efficiency communication systems. The main contributions are as follows: 1. A hybrid message passing algorithm is proposed for faster-than-Nyquist, which solves the problem of joint data detection and channel estimation when the channel coefficients are unknown. To fully exploit the known ISI imposed by FTN signaling, the interference induced by FTN signaling and channel fading are intentionally separated. 2. Gaussian message passing and variational inference based estimation algorithms are proposed for faster-than-Nyquist signaling detection in doubly selective channels. Iterative receivers using mean field and Bethe approximations based on variational inference framework are proposed. Moreover, a novel Gaussian message passing based FTN signaling detection algorithm is proposed. 3. An energy minimisation based SCMA decoding algorithm is proposed and convergence analysis of the proposed algorithm is derived. Following optimisation theory and variational free energy framework, the posterior distribution of data symbol is derived in closed form. Then, the convergence property of the proposed algorithm is analysed. 4. A stretched factor graph is designed for MIMO-SCMA system in order to reduce the receiver complexity. Then, a convergence guaranteed message passing algorithm is proposed by convexifying the Bethe free energy. Finally, cooperative communication methods based on belief consensus and alternative direction method of multipliers are proposed. 5. A low complexity detection algorithm is proposed for faster-than-Nyquist SCMA system, which enables joint channel estimation, decoding and user activity detection in grant-free systems. The combination of FTN signaling with SCMA to further enhance the spectral efficiency is first considered. Then, a merging belief propagation and expectation propagation algorithm is proposed to estimate channel state and perform SCMA decoding.
Please use this identifier to cite or link to this item: