Inverse Design in Distributed Circuits Using Single-Step Reinforcement Learning

TL;DR

This paper introduces DCIDA, a reinforcement learning framework that efficiently explores distributed circuit designs by learning a policy to generate near-optimal configurations, handling complex, non-differentiable evaluation scenarios.

Contribution

DCIDA is a novel single-step reinforcement learning approach that learns a joint sampling policy for distributed circuit design, accommodating complex topologies and evaluation methods.

Findings

DCIDA significantly reduces design error compared to state-of-the-art methods.

The Transformer-based policy network outperforms existing approaches in complex transfer functions.

DCIDA effectively handles non-differentiable evaluation procedures and diverse design topologies.

Abstract

The goal of inverse design in distributed circuits is to generate near-optimal designs that meet a desirable transfer function specification. Existing design exploration methods use some combination of strategies involving artificial grids, differentiable evaluation procedures, and specific template topologies. However, real-world design practices often require non-differentiable evaluation procedures, varying topologies, and near-continuous placement spaces. In this paper, we propose DCIDA, a design exploration framework that learns a near-optimal design sampling policy for a target transfer function. DCIDA decides all design factors in a compound single-step action by sampling from a set of jointly-trained conditional distributions generated by the policy. Utilizing an injective interdependent ``map", DCIDA transforms raw sampled design ``actions" into uniquely equivalent physical representations, enabling the framework to learn the conditional dependencies among joint ``raw'' design decisions. Our experiments demonstrate DCIDA's Transformer-based policy network achieves significant reductions in design error compared to state-of-the-art approaches, with significantly better fit in cases involving more complex transfer functions.