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 $- 35~^{\circ }$ C to $+ 75~^{\circ }$ 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.