Active Constellation Modification Technique for PAPR Reduction of OFDM Signals

The paper proposes an active constellation modification technique to lower peak-to-average power ratio (PAPR) for orthogonal frequency division multiplexing (OFDM) signals. First, the authors introduce some algorithms to reduce computational complexity while ensuring reduced PAPR performance, including the optimal non-iterative ACE algorithm, the suboptimal non-iterative active constellation extension (ACE) algorithm, and the constrained clipping noise filtering (CCNF) algorithm. Next, a hybrid technique and its hardware architecture are introduced that combines the proposed constrained clipping noise filtering algorithm with the proposed suboptimal non-iterative ACE algorithm. This technique aims to improve peak reduction performance while ensuring the overall criteria of the OFDM system. The proposed hybrid technique is then tested through Matlab simulation and FPGA implementation. The MATLAB simulation results allow the evaluation of the effectiveness of lowering PAPR as well as the bit error rate (BER) performance, comparing it with various other methods. Meanwhile, the FPGA’s experimental results evaluate its hardware resource utilization. Specifically, the proposed technique can obtain a reduction of 6.61 dB compared to the original OFDM signal at a CCDF probability level of 10−4, while resulting in an insignificant rise in the signal’s average power of about 0.09 dB. The experimental results also demonstrate the effectiveness of the proposed technique compared to existing ACE and CAF (clipping and filtering) methods in terms of BER efficiency and hardware resource consumption.