Pre-Equalization Algorithms for Nonlinear Phase-Frequency Correction in FI-DAC

The frequency interleaved digital-to-analog converter (FI-DAC) could enhance the bandwidth of the arbitrary waveform generator (AWG). However, it introduces nonlinear phase-frequency errors during multisubband signal stitching. Conventional phase-frequency error calibration methods exhibit limitations in local calibration accuracy, subband-specific correction, and adaptability under varying temperature conditions. To address these issues, this article proposes a Bayesian optimization (BO)-Adabound optimization method by introducing an integral constraint factor. Furthermore, a two-stage Adabound algorithm incorporating Bayesian optimization (BO-DouAdaB) is developed based on Adabound to realize pre-equalizer construction under multiple temperature conditions. In addition, this article completes the calibration of linear delay errors and establishes a comprehensive correction approach that covers phase-frequency errors. The experimental results show that within the test range of -35°C to + 75°C, the BO-DouAdaB optimized pre-equalizer achieves precise calibration performance. These results validate that the proposed methods ensure phase calibration accuracy and enhance the signal generation capability and signal quality of FI-DAC under wide temperature conditions.