Beamforming with 2D-AOA estimation for pilot contamination reduction in massive MIMO

• DOI: 10.1007/s11235-018-0529-y
• Publication Date: 2018-11-23T04:11:33Z
• Authors: Ehab Ali; Mahamod Ismail; Rosdiadee Nordin; Nor Fadzilah Abdulah
• Keywords: Pilot signal; SIGNAL (programming language)

The upcoming fifth-generation wireless mobile systems are projected to be commercialized in the year 2020 and are set to play the main role in massive multiple input-multiple output (mMIMO) integrated with the technologies of beamforming antenna arrays. However, the performance of mMIMO is constrained by inter-cell interference from adjacent cells triggered by the reused pilot, an issue termed pilot contamination. To address this issue, this paper focuses on mitigating pilot contamination by implementing a suboptimal spatial source detection method based on the beamforming approach using a two-dimension-unitary estimation of the signal parameters through rotational invariance techniques algorithm. We jointly used 2D-AOA information (azimuth and elevation angles) and statistical channel estimations at the BSs to identify the channels correlation condition and evaluate the sum rate performance of users based on AOA. The detected signals from the uniform rectangular array are used to segregate the required signal from the interfering signal without any change to the pilot construction of the training signals. The performance of the minimum mean squared error beamforming (MMSE-beamforming) technique in the multi-cell mMIMO system of the aforementioned method is numerically evaluated and then compared with conventional methods that depend only on pilot identity information. The simulation reveals that the achievable sum rate gains of 2D-AOA and the pilot identity information techniques (best case) with respect to deterministic MMSE are 96.3% and 85.4%, respectively, thus showing the prospect of eliminating the majority of pilot contamination using the 2D-AOA estimation.