Fast peak-power reduction for MIMO-OFDM systems with diversity

A robust peak-power reduction technique called active constellation extension (ACE) reduces the peak-to-average power ratio (PAPR) of single-input single-output (SISO) orthogonal frequency-division modulated (OFDM) signals by extending the outer constellation points in a way that minimizes the PAPR of an OFDM transmit signal, without reducing the bandwidth or increasing bit error rate (BER). Recent work includes extensions of the concept of ACE using a modi ed smart gradient-project algorithm for MIMO-OFDM systems. We extend the e cient ACE smart gradient-project (SGP) method to space-time block coded (STBC), space-frequency block coded (SFBC) and V-BLAST OFDM systems. The proposed peak-power reduction method can be applied to any space-time-frequency (STF) block code, and its performance is bounded by the code structure. Simulations show PAPR reduction gains of 4.19 and 3.57 dB under QPSK for Alamouti STBC and SFBC, respectively. In addition, we extend the e cient ACE-SGP method to rotated constellations for SISO-OFDM systems, an alternative method recently invented for providing diversity. Simulation results show approximately 3.46 dB and 2.08 dB of PAPR reduction for QPSK and 16-QAM original constellations, respectively, for 256 subcarriers. Results also show that 4.29 dB and 3.41 dB of PAR reduction are obtained for QPSK and 16-QAM original constellations, respectively, for 1024 subcarriers. Furthermore, we show how the PAPR reduction method known as tone injection can be a very practical transmission scheme for coded-OFDM systems, with a slight compromise of the BER performance. This is accomplished by developing e cient, near-optimal, convolutional and LDPC decoders for the extended non-bijective constellations introduced by tone injection. Simulation results for convolutional and LDPC coded-OFDM systems show BER performance comparable to conventional coded OFDM with greatly reduced PAPR. Furthermore, results show that the PAPR performance obtained is very close to the 64-QAM single-carrier peak-power performance. So, the peak-power problem is solved for large constellations