Font Size: a A A

Research On Pilot Contamination Of Massive MIMO System

Posted on:2018-07-24Degree:MasterType:Thesis
Country:ChinaCandidate:D D MengFull Text:PDF
GTID:2348330518493321Subject:Information and Communication Engineering
Abstract/Summary:PDF Full Text Request
Studies have shown that, Massive Multi-Input Multi-Output (Mas-sive MIMO) system has many advantages like BS can serve many users in the same range of spectrum band with low transmit power and obtain higher transmission rate. The highly increase number of antennas at BS make the channel vectors orthogonal, then some simple detect algorithm can obtain optimal performance which is important for complex structure of Massive MIMO system. But the improvement of performance of Mas-sive MIMO system is based on the perfect channel state information.When all cells share same pilots, the channel estimation of one user will be interfered by the channel information of the users in other cells with the same pilots assigned to them, thus inaccuracy channel estimation causes pilot contamination (PC) and limits the improvement of system performanceMassive MIMO system model is introduced, and the reason of PC and the impact are analysed in this paper. The main methods to reduce the PC at present, such as channel estimation, pilot sequences design and as-signment, multi-cell precoding and so on, are introduced. The paper will mainly propose a pilot design scheme for user distinguished Massive MIMO system and DL multi-cell precoding method next. For now, most researches about reducing PC are treating users in the same way and the user experience is not considered in real communication system. This de-sign scheme divides users within a cell into center user group and edge user group according to large scale fading factor. For edge user group, the pilot vectors are orthogonal to insure the channel estimation accuracy of edge users, which can reduce the further interference caused by PC. For center user group, however, different pilot groups with different phase factors are assigned to different cells, which can reduce the correlation of different pilot vectors assigned to different users in different cells and further reduce the PC of whole system. Among the whole process, the se-lection of phase factors related to different cells is the key work. The range of the phase factor can be derived by obeying the rule of all the pi-lot vectors within a cell must be orthogonal. After the range is obtained,parts of phase factors can be assigned to different cells.The non-reused pilot design method not only improves the quality of service of edge users, but reduces the PC of center users and improves performance of system. The multi-cell block diagonal (BD) precoding is adopted in DL communication to reduce the interference one user added to another in other cell which can obtain more throughput.
Keywords/Search Tags:Massive MIMO, pilot contamination, pilot design, multi-cell precoding
PDF Full Text Request
Related items