SISO Digital Predistortion for Concurrent Dual-Band Power Amplifier Using Baseband Stitching Technique

In this paper, a single input and single output (SISO) digital predistortion architecture is proposed to linearize the concurrent dual-band power amplifier. The signals used for training the predistorter are obtained by baseband stitching technique. To predict the memory effect precisely, a simplified baseband equivalent Volterra model with dynamic memory depth is employed. The performance of the proposed method is evaluated by linearizing a concurrent dual-band power amplifier which works at 3.3GHz and 3.8GHz. Compared to the conventional 2D predistorter, experiment results show that the proposed method can reduce the number of model coefficients without sacrificing performance.