A Digital Correction Method to Reduce the Nonlinearity of Terahertz Transmitters

Shanghui Xiao,Chao Fan,Xiangjie Xia,Jian Liu,Xin Quan,Ying Liu,Shihai Shao,Youxi Tang
DOI: https://doi.org/10.1109/gcwkshps52748.2021.9682072
2021-01-01
Abstract:Terahertz frequency ranging from 100 GHz to 10 THz is able to provide sufficient bandwidth and has gained much attention in the area of 6G wireless communications. However, due to the inherent nonlinearity of terahertz components and integrated modules, the current modulation schemes are usually limited to simple schemes with low spectrum efficiency, such as QPSK. To improve the transmission spectrum efficiency, in this paper, a digital correction method is proposed to alleviate the terahertz transmitter nonlinearity at the receiver side. In this method, a simplified nonlinearity correction model is first constructed at the symbol level and then trained by using some known transmitted symbols. Subsequently, the trained model is inserted into the receiver link to compensate for the nonlinear distortion of the transmitter to improve the quality of the received symbols. As a result, modulation schemes with higher spectrum efficiency, such as 64-QAM, can be adopted to improve the transmission speed within the limited bandwidth constrained by practical terahertz components. Simulations verify that the signal quality can be effectively improved in terms of the error vector magnitude (EVM).
What problem does this paper attempt to address?