Quasi-Optimum Detection of OFDM with Cartesian Nonlinearities

In the last years, it has been shown that nonlinear (NL) orthogonal frequency division multiplexing (OFDM) can outperform linear OFDM because the nonlinear distortion has useful information about the transmitted signals. However, only a maximum likelihood (ML) receiver can exploit this information. Despite its potential, the optimal ML receiver for NL OFDM is highly complex, and its performance is difficult to simulate. Moreover, although some theoretical performance bounds exist for OFDM transmissions involving many subcarriers and high signal-to-noise ratio (SNR), the behavior of NL OFDM under more practical SNR conditions remains insufficiently explored. This paper focuses on optimal detection methods for NL OFDM systems. We present a comprehensive analysis of how the distortion received on different subcarriers contributes to the signal of a specific subcarrier and derive a performance bound for the ML detection of NL OFDM that is applicable across a broad range of SNR values. Furthermore, we propose a practical iterative decision-directed receiver that achieves significant performance improvements over linear OFDM in both uncoded and coded setups.

Quasi-Optimum Detection of OFDM with Cartesian Nonlinearities

Daniel Dinis, IST-Universidade de Lisboa, Aalto University, Copelabs; Diogo Costa, IST, Instituto Superior Técnico; Aalto University; João Guerreiro, FCT-Universidade Nova de Lisboa, Instituto de Telecomunicações; Marko Beko, Instituto Superior Técnico, Universidade de Lisboa/COPELABS; Risto Wichman, Aalto University