In recent years, fully-connected hybrid beamforming (FC-HBF) architecture has aroused widespread interest for millimeter wave (mmWave) massive multi-input multi-output (MIMO) communication systems. However, the FC-HBF structure suffers from significant linearity deterioration, limiting its applications in actual mmWave transmitters. To resolve this issue, an effective digital predistortion (DPD) method utilizing analytical multi-input behavioral models is proposed in this paper for linearizing the FC-HBF system. Based on the nonlinearity analysis and behavioral modeling of the array response, three analytical multi-input models are derived by embedding the priori beamforming information in the predistorter. The complexity of the proposed analytical models is significantly reduced compared to the state-of-the-art. Numerical simulations and experimental measurement are carried out on a$4\times 64$mmWave FC-HBF array and$2\times 16$quasi-test platform respectively to validate the performance of the proposed DPDs against the state-of-the-art DPDs, which show significant linearization abilities to compensate for the nonlinear distortions of beam signals. The proof-of-the-concept validations in this paper indicate that the proposed scheme is fully capable of linearizing an mmWave FC-HBF array.

Nowadays, various fonts are applied in many fields, and the generation of multiple fonts by computer plays an important role in the inheritance, development and innovation of Chinese culture. Aiming at the existing font generation methods, which have some problems such as stroke deletion, artifact and blur, this paper proposes Chinese font translation with improved wasserstein generative adversarial network. The wasserstein distance is used to measure distance and difference between the two distributions, and the gradient penalty mechanism is used instead of the weight clipping. The residual dense blocks with better flexibility are selected as the core component to extract the features fully and enhance the information transmission between network layers. It realizes the style migration between different Chinese fonts. Experiments show that the proposed method has better performance in font generation details, simplifies the font translation process, and improves the fidelity of font generation.