Fast Response Prediction Method Based on Bidirectional Long Short-Term Memory for High-Speed Links

Abstract

There is a recognized need to evaluate the performance of high-speed links accurately and quickly with increasing clock frequency. To evaluate the performance of the systems, the transient simulation method and the fast time-domain simulation method are adopted. Unfortunately, neither method can guarantee both the efficiency and accuracy of the results at the same time. In this article, a modeling method that can overcome the shortcomings of both the transient simulation method and the fast time-domain simulation method is proposed for obtaining responses of high-speed links based on bidirectional long short-term memory (Bi-LSTM). Here, the pulse response and response of a certain number of bits are first simulated or measured, which are used to determine training datasets. Then, the Bi-LSTM-based model uses the generated training datasets to build a precise model. Next, the developed model predicts the response of the high-speed link. Compared to the transient simulation method and fast time-domain simulation method, the results show that the proposed method can consider both accuracy and efficiency. In addition, 33 simulation experiments are performed to demonstrate the robustness of the proposed method. Finally, the accuracy of the proposed method is validated by comparing the measured and predicted results of the high-speed links with different nonlinearities.