Precompensation and System Parameters Estimation for Low-Cost Nonlinear Tera-Hertz Transmitters in the Presence of I/Q Imbalance

Tera-Hertz (THz) transmission can offer several attractive applications, yet developing low-cost energy-efficient THz devices is at an early stage. The most promising low-cost THz transmitter architecture in the literature is the so-called frequency-multiplier-last architecture. However, it is incapable of transmitting quadrature amplitude modulation (QAM) due to the architecture's inherent nonlinear distortions. We study such nonlinear THz communication systems by incorporating the nonlinearity aspects of the low-cost THz devices and the inphase and quadrature (I/Q) imbalance effect into the signal model. Then, we propose a precompensation scheme to compensate the nonlinearity and I/Q imbalance effects, thus enabling the QAM-capable frequency-multiplier-last architecture for THz systems. The proposed precompensation scheme requires the knowledge of the system parameters. To estimate the system parameters, we propose a maximum-likelihood estimator and its practical implementation via an alternating estimation algorithm. We also derive closed-form expressions for the Cramér-Rao lower bounds (CRLBs) of the system parameters estimation, and design the pilot sequence used in estimating the system parameters. Numerical results show that the proposed precompensation schemes overcome the prominent problems experienced in the existing THz systems, namely severe nonlinear distortions of the modulation symbols as well as spectral spreading and/or large spectrum sidelobes, and mitigate the I/Q imbalance effect.