Files in this item
|(no description provided)|
|Title:||Automatic test generation for bit-serial VLSI digital signal processors|
|Author(s):||Roy, Rabindra Kumar|
|Doctoral Committee Chair(s):||Patel, Janak H.|
|Department / Program:||Electrical and Computer Engineering|
|Degree Granting Institution:||University of Illinois at Urbana-Champaign|
|Subject(s):||Engineering, Electronics and Electrical|
|Abstract:||Linear digital signal processors, commonly implemented using silicon compilers in bit-serial architecture, are very difficult to test for manufacturing defects due to deep sequentiality, low controllability and observability, and high latency. A novel hierarchical testing approach has been proposed, integrating three automatic test generators. FEAST (Functional Extractor and Sequential Test generator) operates at the high level, where the circuit is described as an interconnection of arithmetic modules. SEAT (Sequential Array Test generator) and CREST (Constrained Sequential Test generator) operate at the low level description of the individual modules.
FEAST extracts various matrices corresponding to the state space description of the circuit, and the functional constraints that must be satisfied the inputs of the module under test (MUT). The constraints and the description of the MUT are passed to CREST, which generates tests in the binary domain. FEAST translates these tests into the numerical domain to find the numbers to be applied at the primary inputs to achieve the tests at the inputs of the MUT. Since the binary representation of these numbers are inexact due to truncation and roundoff, a novel search-based correction scheme is employed to get the exact solution in binary.
SEAT operates on array structures and attempts to generate repetitive tests to apply comprehensive checking experiments to finite state machines within individual cells of the array. SEAT has been very successful on one-dimensional sequential arrays and outperformed an efficient automatic gate-level test generator.
The performance of FEAST and CREST was compared with two other approaches, namely automatic test generation at gate level and random test pattern generation for several large circuits. The new techniques recorded the best combination of all of the performance metrics, namely, fault coverage, test efficiency, test length, and computational time.
|Rights Information:||Copyright 1992 Roy, Rabindra Kumar|
|Date Available in IDEALS:||2011-05-07|
|Identifier in Online Catalog:||AAI9236586|
This item appears in the following Collection(s)
Graduate Dissertations and Theses at Illinois
Graduate Theses and Dissertations at Illinois
Dissertations and Theses - Electrical and Computer Engineering
Dissertations and Theses in Electrical and Computer Engineering