This paper proposes a pilot power allocation scheme that simultaneously reduces both the pilot contamination effect and per-user pilot transmission power for the user-centric cell-free (CF) massive multiple-input multiple-output (mMIMO) systems during the uplink training phase. The proposed pilot power allocation scheme comprises a simple algorithm that tries to minimize the channel estimation error iteratively for improving the estimation quality. It is found from the simulated results that the average convergence time of the proposed scheme demonstrates a scalable attribute. Moreover, numerical results verify the effectiveness of the proposed scheme in improving the 95%-likely spectral efficiency performance by up to 16% and reducing the average pilot transmission power by up to 78 % compared to a scheme that transmits full pilot power (i.e., no control) and a recent pilot power allocation scheme that uses successive approximation technique with first-order Taylor approximation.
Manobendu SarkerAbraham O. Fapojuwo
Manobendu SarkerAbraham O. Fapojuwo
Tong Van LuyenNguyen Van Cuong
Manobendu SarkerAbraham O. Fapojuwo