الفهرس 
Overview of Massive MIMO
Pilot contamination Mitigation TechniquesOnly 14 pages are availabe for public view
 |  | from | 168 |  |  |
| | | from | 168 | | |
Abstract
The wireless data traffic has grown exponentially due to the dramatic increase of con- nected devices and the proliferation of new applications requiring high data rate. Ac- cording to Martin Cooper law, the demand for wireless data traffic will be doubled every eighteen months. Hence, fifth generation (5G) wireless communication networks are re- quired along with new technologies in which many users can be simultaneously served with very high throughput.
One of the promising technologies that can allow 5G cellular networks to meet these demands is the massive Multi-Input Multi-Output (MIMO) technique, where each base station (BS) is equipped with a large number of antennas to serve a multiple number of single-antenna user equipments (UEs), simultaneously in the same time and frequency resource. Massive MIMO systems can substantially enhance the spectral efficiency in order of magnitude compared to the conventional MIMO systems.
Within the cell, the BS is only required to know the channel state information (CSI) in order to do multi-user precoding in the downlink and data detection in the uplink. The uplink CSI is estimated at the BS based on the uplink orthogonal pilot sequences. Because the number of pilot sequences is limited by the channel coherence interval, the pilot sequences are mutually orthogonal in the same cell but necessarily reused in some adjacent cells. Therefore, the estimated CSI at the BS in a given cell is contaminated by the users that use the same pilot sequences in other cells. This effect is called pilot contamination that limits the performance of massive MIMO systems.
In this thesis, we propose two novel pilot allocation schemes in order to mitigate the pilot contamination effect and to enhance the performance of massive MIMO systems
viii
ix
with low complexity of pilot allocation. In these proposed schemes, the cell coverage is partitioned into center and edge zones. Furthermore, the edge zone is sectorized into a number of uniform sectors by the smart antenna technique. The achievable spectral efficiency adopting both the maximum-ratio combining (MRC) and zero-forcing (ZF) schemes are obtained. Then, the obtained spectral efficiency of these proposed schemes are formulated as an optimization problem with an objective of maximizing the overall spectral efficiency over the ratio of center zone radius to the cell radius and the number of scheduled users per cell. After that, this optimization problem is mathematically solved in order to find the optimal solutions. In addition, the upper bound for ergodic capacity is analytically derived in order to illustrate that the proposed schemes are more close to the upper bound than the conventional schemes in the literature.
Finally, the performance of these proposed schemes is investigated through simulations results in which the optimization problem is also solved using the exhaustive search in order to find the optimized spectral efficiency for the proposed schemes. Moreover, the obtained results are compared against the conventional schemes, and the upper bound on the ergodic capacity. It turns out that the spectral efficiency of the proposed schemes always outweigh the spectral efficiency of the conventional schemes along the number of BS antennas and also the performance gap between the upper bound on ergodic capacity and the proposed schemes is less than the performance gap between the upper bound on ergodic capacity and the conventional schemes. This is a consequence of wisely allocating the pilot sequences among the users due to dividing the cell coverage into center and edge zones and sectorizing the edge zone.