publication . Conference object . 2018

A Semi-Formal Technique to Generate Effective Test Sequences for Reconfigurable Scan Networks

Riccardo Cantoro; Aleksa Damljanovic; Matteo Sonza Reorda; Giovanni Squillero;
Open Access English
  • Published: 13 Sep 2018
  • Publisher: Zenodo
  • Country: Italy
Abstract
The broad need to efficiently access all the instrumentation embedded within a semiconductor device called for a standardization, and the reconfigurable scan networks proposed in IEEE 1687 have been demonstrated effective in handling complex infrastructures. At the same time, different techniques have been proposed to test the new circuitry required; however, most of the automatic approaches are either too computationally demanding to be applied in complex cases, or too approximate to yield high-quality tests. This paper models the state of a reconfigurable scan network with a finite state automaton, using the length of the active path as the output alphabet and the configurations as input symbols. Permanent faults are represented as incorrect transitions, and a greedy algorithm is used to generate a functional test sequence able to detect all these multiple state-transition faults. The automaton's state set and the input alphabet are small subsets of the possible ones, and are carefully chosen. Experimental results on ITC'16 benchmarks demonstrate that the proposed approach is broadly applicable; the test sequences are more efficient than the ones previously generated by search heuristics.
Persistent Identifiers
Subjects
free text keywords: State (computer science), Set (abstract data type), Path (graph theory), Finite-state machine, Computer science, Instrumentation (computer programming), Algorithm, Greedy algorithm, Automaton, Heuristics
Related Organizations
Funded by
EC| RESCUE
Project
RESCUE
Interdependent Challenges of Reliability, Security and Quality in Nanoelectronic Systems Design
  • Funder: European Commission (EC)
  • Project Code: 722325
  • Funding stream: H2020 | MSCA-ITN-ETN
Validated by funder
Download fromView all 5 versions
Open Access
ZENODO
Conference object . 2018
Providers: ZENODO
Any information missing or wrong?Report an Issue