Efficient LLR Approximation for Coded Constant Envelope OFDM

Jiabin Tan,Yue Xiao,Lilin Dan,Wei Xiang
DOI: https://doi.org/10.1109/TVT.2022.3231912
IF: 6.8
2023-01-01
IEEE Transactions on Vehicular Technology
Abstract:Constant envelope orthogonal frequency division multiplexing (CE-OFDM) is a promising technique to address the problem of high peak-to-average power ratio (PAPR) by embedding information into the phases of carriers since it was proposed. Meanwhile, in order to further enhance the bit error rate (BER) performance when using forward error correction (FEC) coding, the soft decision should be employed in CE-OFDM systems. However, due to the nonlinear noise in the phase of CE-OFDM, inaccurate log-likelihood ratios (LLRs) cause degradation of the decoding performance. To alleviate this problem, the systematic cause analysis of the phase noise is conducted followed by an efficient phase-noise oriented (PNO) algorithm for the phase demodulator in the CE-OFDM system. We further propose another stationary region Max-log-MAP (SR-MLM) algorithm with the phase unwrapper towards low implementation costs. Simulation results demonstrate that the proposed LLR approximation algorithms are capable of improving the BER performance of coded CE-OFDM compared to its conventional counterpart.
What problem does this paper attempt to address?