Home // CENICS 2016, The Ninth International Conference on Advances in Circuits, Electronics and Micro-electronics // View article

A Hotspot Detection Method Based on Approximate String Search

Authors:
Shuma Tamagawa
Ryo Fujimoto
Masato Inagi
Shinobu Nagayama
Shin'ichi Wakabayashi

Keywords: lithography; hotspot; optical simulation; approximate string matching

Abstract:
In this paper, we propose an approximate string search-based method for detecting hotspots on mask patterns used in very-large-scale integration (VLSI). In mask patterns for manufacturing VLSI chips, there are some local patterns which induce open/short circuits and thus failures. Such patterns are called hotspots. They are detected by optical simulation before manufacturing. Since it requires very long time, it is desirable to detect hotspot candidates in order to limit the target regions of optical simulation. For the detection, methods which check matching between each pattern from a pre-defined hotspot library and the mask pattern are attracting attention. At first, our proposed method transforms the mask pattern and the pre-defined hotspot patterns, which are two-dimensional, into one-dimensional strings. Then, it finds hotspot candidates by using approximate string search to detect patterns similar to hotspot patterns. The transformation is performed in a particular way to efficiently realize quasi-two-dimensional search by using string search. In addition, we focus on the distance between wires as a metric to find hotspot candidates, and give a priority to patterns which have wires with a shorter distance. To evaluate the effectiveness of our method, we conducted some experiments.

Pages: 6 to 12

Copyright: Copyright (c) IARIA, 2016

Publication date: July 24, 2016

Published in: conference

ISSN: 2308-426X

ISBN: 978-1-61208-496-1

Location: Nice, France

Dates: from July 24, 2016 to July 28, 2016