Skip to content
STIMSMITH

Automatic Constraint Extraction

Concept WIKI v1 · 5/30/2026

Automatic constraint extraction is a verification concept described in the context of coverage-directed test selection: instead of relying only on manually written constraints to steer constrained-random tests, a supervised-learning method uses coverage feedback to prioritize randomly generated tests that are likely to improve functional coverage.

Overview

Automatic constraint extraction is described as part of a method for improving simulation-based hardware verification. In constrained-random test generation, tests are randomized to increase diversity, but many tests can repeatedly exercise the same design logic. Constraints are typically written manually to bias random tests toward interesting, hard-to-reach, or not-yet-tested logic. [C1]

The cited work introduces a method for automatic constraint extraction and test selection, called coverage-directed test selection. The method is based on supervised learning from coverage feedback. [C2]

Motivation

As verification progresses, many constrained-random tests may contribute little or nothing to functional coverage. When generating stimuli is cheaper than simulation, the paper proposes generating many random tests first, selecting the most effective subset, and simulating only that subset. [C3]

Role in coverage-directed test selection

In the described approach, supervised learning is used to bias selection toward tests that have a high probability of increasing functional coverage. These tests are then prioritized for simulation. [C4]

Reported benefits

The paper reports that coverage-directed test selection can reduce manual constraint writing, prioritize effective tests, reduce verification resource consumption, and accelerate coverage closure on a large, real-life industrial hardware design. [C5]

Limitations of available evidence

The provided evidence is the paper abstract and submission metadata. It supports the high-level definition, motivation, and reported benefits, but does not provide algorithmic details of how constraints are extracted internally.

LINKED ENTITIES

4 links
Coverage-Directed Test Selection implemented_as_or_part_of
constrained-random test generation applies_to
simulation-based verification used_in
Functional Coverage optimizes_for

CITATIONS

5 sources
5 citations
[1] Constrained-random test generation uses randomness for diversity, but tests may repeatedly exercise the same design logic; constraints are typically written manually to bias tests toward interesting, hard-to-reach, or untested logic. Supervised Learning for Coverage-Directed Test Selection in Simulation-Based Verification
[2] The paper introduces a method for automatic constraint extraction and test selection called coverage-directed test selection, based on supervised learning from coverage feedback. Supervised Learning for Coverage-Directed Test Selection in Simulation-Based Verification
[3] As verification progresses, many constrained-random tests yield little or no functional-coverage effect, motivating generation of many random tests followed by selection and simulation of only the most effective subset. Supervised Learning for Coverage-Directed Test Selection in Simulation-Based Verification
[4] Coverage-directed test selection biases selection toward tests with a high probability of increasing functional coverage and prioritizes them for simulation. Supervised Learning for Coverage-Directed Test Selection in Simulation-Based Verification
[5] The paper reports that coverage-directed test selection can reduce manual constraint writing, prioritize effective tests, reduce verification resource consumption, and accelerate coverage closure on a large real-life industrial hardware design. Supervised Learning for Coverage-Directed Test Selection in Simulation-Based Verification