Pilot Resource Management for TDD Massive MIMO Systems

• Main Authors: Yun-KueiHua, 華雲貴
• Other Authors: Wenson Chang
• Format: Others
• Language: en_US
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/5uvgh3

博士 === 國立成功大學 === 電腦與通信工程研究所 === 107 === This dissertation has four parts. The first part we focus on a partial-time-shifted pilot (PTSP) scheme to find a better tradeoff between the user accommodation and orthogonality of pilots for the massive multiple-input multiple-output (MIMO) system. To this end, the pilot symbols rather than the pilot sequences are applied. Also, the neighboring cells are allowed to transmit the pilot signals using the overlapped symbol periods. Despite the stronger interference, the larger user accommodation as well as the near-far effect incurred by using the pilot symbols can contribute to the higher sum transmission rates for both the uplink and downlink cases. Second, we focuse on a cocktail approach, to delicately integrate the two conventional countermeasures, i.e., the fractional pilot reuse and time-shifted pilots (TSP) schemes, named it dual partial-time-shifted pilot (DTSP) scheme, for the pilot contamination problem in the full-duplex massive multiple-input multiple-output system. Specifically, the TSP scheme is applied to duplicate for the inner and edge regions of a cell such that the number of accommodated users can be doubled. To achieve this, the base station should operate in the full-duplex mode; also, the pilot symbols rather than the pilot sequences are used to attain the user diversity gain. The simulation results verify the accuracy of the numerical analysis and the remarkably boosted sum rate for both the downlink and uplink cases. For the third part, we develop a general multi-ring TSP (MR-TSP) scheme by dividing a cell into multiple rings; and allocating non-overlapped timeslots to different rings for transmitting pilots. Moreover, similar to the conventional TSP scheme, the same time slots can still be shared with cells belonging to different groups. Based on the analytical and simulation results, it is found that three rings are recommended for the TSP-based pilot contamination schemes. Finally, we develop the multi-cell cooperative scheduling (MCCS) algorithm to properly arrange users for completely reusing the limited orthogonal pilots among the neighboring cells so that the impact of pilot contamination (PC) in the massive MIMO systems can be alleviated. To this end, the multi-cell cooperative scheduling indexes (COSIs) are defined for maximizing the data rate, maximizing the Jain's fairness index and reach a better tradeoff in-between (which are denoted by CMDR, CMMF and CPF COSIs, respectively).