Metamodel-Assisted Fast and Accurate Optimization of an OP-AMP for Biomedical Applications

Abstract

The optimized OP-AMPs resulting out of a traditional flows, although may meet the given specifications after consuming significant design cycle time, do not guarantee an optimal system performance. In this paper, a three-step polynomial metamodel-assisted OP-AMP optimization flow is proposed to address these issues. The flow incorporate polynomial metamodeling, Verilog-AMS integration, and a customized Cuckoo Search optimization. To the best of the authors' knowledge, this paper for the first time presents such a design flow for state-of-the art OP-AMP optimization. Highly accurate and ultra-fast (~17000× speedup compared to traditional methods) polynomial metamodels are generated to estimate OP-AMP performance. An OP-AMP meta-macro model is constructed and is integrated into a Verilog-AMS module (called Verilog-AMS-POM) to facilitate fast time-domain simulations. The core optimization module is a customized Cuckoo Search algorithm which produces promising optimized results. The OP-AMP power dissipation is reduced from 252.8 μW to 65.5 μW (3.86× improvement).