SCA Power optimization in Cell-Free Massive MIMO with Short-Term Power Constraints

Abstract

In this paper, the performance of downlink cell-free massive multiple-input multiple-output (mMIMO) system with short-term power constraints is investigated. To maximize the sum downlink user signal-interference-noise-ratio (SINR), this paper proposes a sequential convex approximation (SCA) power optimization algorithm for cell-free mMIMO with normalized conjugate beamforming (CB) precoder. This proposed SCA algorithm is proved to be a second order cone programming (SOCP). Then, under the condition that the achievable data SINR is greater than or equal to a certain quality of service (QoS) threshold, a new sum power minimization algorithm is presented. It also can be reformulated as a SOCP. Numerical results indicate that the achievable sum downlink SINR monotonically increases at each iteration and the sum SINR tends to a stable value after 7 iterations. In addition, the power minimization algorithm works well in satisfying target users’ SINR constraints and minimizing the sum downlink transmission power.