Low-Complexity PAPR Minimization for Symbol Level Precoded Multi-User MISO-OFDM System

This letter proposes a method exploiting constructive interference (CI) to reduce the transmit signal's peak-to-average power ratio (PAPR), while keeping the total transmission power as low as possible. An optimization problem that jointly performs power minimization and PAPR reduction is formulated, which is however difficult to solve directly due to the non-convex PAPR constraint. To obtain a feasible solution in low complexity, by using the vectorization method and introducing a regularization factor, we relax the PAPR constraint. The original optimization problem is transformed into a convex problem that can be solved with an improved fast iterative shrinkage-thresholding algorithm (FISTA). Numerical results are presented to show 1dB savings in terms of transmission power and 52% savings in terms of PAPR compared with state-of-the-art PAPR minimization techniques.