Abstract

In recent years, massive Multi-User Multi-Input Multi-Output (MU-MIMO) system has attracted significant research interests in mobile communication systems. It has been considered as one of the promising technologies for 5G mobile wireless networks. In massive MU-MIMO system, the base station (BS) is equipped with a very large number of antenna elements and simultaneously serves a large number of single-antenna users. Compared to traditional MIMO system with fewer antennas, massive MU-MIMO system can offer many advantages such as significant improvements in both spectral and power efficiencies. However, the channel estimation in massive MU-MIMO system is particularly challenging due to large number of channel matrix entries to be estimated within a limited coherence time interval. This problem occurs in a single-cell case where both dimensions of the channel matrix grow large. Also, It happens in the multi-cell setting due to the pilot contamination effect.
In this thesis, the problem of channel estimation in both single-cell and multi-cell time division duplex (TDD) massive MU-MIMO systems is studied. Thus, two-channel estimation namely “nuclear norm (NN)” and “iterative weighted nuclear norm (IWNN)” approximation techniques are proposed to solve the channel estimation problem in both systems.
First, channel estimation in a single-cell TDD massive MU-MIMO system is formulated as a convex nuclear norm optimization problem with regularization parameter γ. In this study, the regularization parameter γ is selected based on the cross-validation (CV) curve method. The simulation results in terms of the normalized mean square error (NMSE) and uplink achievable sum-rate (ASR) are provided to show the effectiveness of the NN proposed scheme compared to the conventional least square (LS) estimator. Then, the IWNN approximation is proposed to improve the performance of the NN method. Thus, the channel estimation in a single-cell TDD massive MU-MIMO system is formulated as a weighted nuclear norm optimization problem. The simulation results show the effectiveness of the IWNN estimation approach compared to the standard NN and conventional LS estimation methods in terms of the NMSE and ASR.
Second, both previous estimation techniques are extended to apply in a multi-cell TDD massive MU-MIMO system to mitigate pilot contamination effect. The simulation results in terms of the NMSE and uplink ASR show that the IWNN scheme outperforms the NN and LS estimations in the presence of high pilot contamination effect.
Finally, a novel channel estimation scheme namely “Approximate minimum mean square error (AMMSE)” is proposed to reduce the computational complexity of the minimum mean square error (MMSE) estimator which was proposed for multi-cell TDD massive MU-MIMO system. Furthermore, a brief analysis of the computational complexity regarding the number of multiplications of the proposed AMMSE estimator is provided. It has been shown that the complexity of the proposed AMMSE estimator is reduced compared to the conventional MMSE estimator. The simulation results in terms of the NMSE and the uplink ASR performances show the proposed AMMSE estimation performance is almost the same as the conventional MMSE estimator under two different scenarios: noise-limited and pilot contamination.