CASE-BASED TIMING ASSESSMENT TOOL

Period of Performance: 01/01/1992 - 12/31/1992

$485K

Phase 2 SBIR

Recipient Firm

Advanced System Technologies, Inc.
12200 E. Briarwood Avenue, Suite 260
Englewood, CO 80112

Abstract

THE MOST CRITICAL DEFICIENCY WITH CASE PRODUCTS AS APPLIED TO TIME-CRITICAL SOFTWARE DESIGN IS THEIR INABILITY TO ADEQUATELY SUPPORT THE TRADE-OFF PROCESS INVOLVING TIMING AND PERFORMANCE ISSUES. ONLY A FEW OF THE SO-CALLED REAL-TIME TOOLS SUPPORT THE ALLOCATION AND BUDGETING OF TIMING REQUIREMENTS. NONE OF THE AVAILABLE CASE TOOLS PROVIDE THE USER WITH ESTIMATES OF DELIVERED TIMING THAT ARE DERIVED FROM THE SYSTEM DESIGN AND THE WORK-LOAD FACTORS. QUANTITATIVE FEEDBACK REGARDING THE IMPACT OF SOFTWARE DESIGN DECISIONS ON SYSTEM TIMING IS CRITICAL TO SYSTEM SUCCESS, SINCE VIRTUALLY EVERY STEP IN THE DESIGN PROCESS CAN HAVE AN EFFECT ON PERFORMANCE. THIS PROJECT ADDRESSES A SOLUTION TO THIS DEFICIENCY BY INTEGRATING STATE-OF-THE-ART PERFORMANCE MODELING TECHNIQUES WITH CASE TOOL SOFTWARE-DESIGN MODELS. THE TECHNICAL OBJECTIVES OF PHASE I INCLUDE FORMALIZING A STIMULUS-RESPONSE (S-R) DIAGRAM REPRESENTATION WHICH IS ABLE TO SUPPORT DETAILED TIMING SIMULATIONS AND DETERMINING COMPATIBILITY BETWEEN S-R REPRESENTATION-AND STATE TRANSITION-BASED DESIGN REPRESENTATIONS. STATE-TRANSITION-BASED DESIGN REPRESENTATIONS (WARD AND MELLOR, HATLEY, ESML, STATECHARTS, ETC.) ARE BEING SPECIFICALLY TARGETED BECAUSE OF THEIR PREDOMINANCE IN THE CASE USER COMMUNITY. THE MOST CRITICAL DEFICIENCY WITH CASE PRODUCTS AS APPLIED TO TIME-CRITICAL SOFTWARE DESIGN IS THEIR INABILITY TO ADEQUATELY SUPPORT THE TRADE-OFF PROCESS INVOLVING TIMING AND PERFORMANCE ISSUES. ONLY A FEW OF THE SO-CALLED REAL-TIME TOOLS SUPPORT THE ALLOCATION AND BUDGETING OF TIMING REQUIREMENTS. NONE OF THE AVAILABLE CASE TOOLS PROVIDE THE USER WITH ESTIMATES OF DELIVERED TIMING THAT ARE DERIVED FROM THE SYSTEM DESIGN AND THE WORK-LOAD FACTORS. QUANTITATIVE FEEDBACK REGARDING THE IMPACT OF SOFTWARE DESIGN DECISIONS ON SYSTEM TIMING IS CRITICAL TO SYSTEM SUCCESS, SINCE VIRTUALLY EVERY STEP IN THE DESIGN PROCESS CAN HAVE AN EFFECT ON PERFORMANCE. THIS PROJECT ADDRESSES A SOLUTION TO THIS DEFICIENCY BY INTEGRATING STATE-OF-THE-ART PERFORMANCE MODELING TECHNIQUES WITH CASE TOOL SOFTWARE-DESIGN MODELS. THE TECHNICAL OBJECTIVES OF PHASE I INCLUDE FORMALIZING A STIMULUS-RESPONSE (S-R) DIAGRAM REPRESENTATION WHICH IS ABLE TO SUPPORT DETAILED TIMING SIMULATIONS AND DETERMINING COMPATIBILITY BETWEEN S-R REPRESENTATION-AND STATE TRANSITION-BASED DESIGN REPRESENTATIONS. STATE-TRANSITION-BASED DESIGN REPRESENTATIONS (WARD AND MELLOR, HATLEY, ESML, STATECHARTS, ETC.) ARE BEING SPECIFICALLY TARGETED BECAUSE OF THEIR PREDOMINANCE IN THE CASE USER COMMUNITY.