Constraint manipulation techniques for synthesis and verification of embedded systems.

摘要

Recently, the industry has adopted the programmable system-on-chip platform as architecture of choice that meets the performance and feature demands posed by many multi-application usercustomizable embedded systems. Due to increased design complexities, shortened time-to-market, and high sensitivity of product's economic success to R&D costs, system designers have adopted design reuse as a standard hardware and software design paradigm. In this work, we present several design methodologies that use constraint manipulation as means of (i) enabling effective debugging of core-based systems-on-chip as well as (ii) enforcing protection of intellectual property (IP) from fraud and misappropriation.; To facilitate system debugging, we introduce a cut-based functional debugging paradigm that leverages on the advantages of both emulation and simulation. The approach enables the user to run test sequences in emulation for fast execution and upon error detection, roll-back to an arbitrary instance in execution time, and transparently switch to simulation-based debugging for full design visibility and controllability.; We have developed a symbolic debugger, a tool that enables designers to interact with the executing code at the source level. When a behavioral specification is optimized using transformations, values of source code variables may be either inaccessible at run-time or inconsistent with what the designer expects. In response to a query, the symbolic debugger retrieves and displays values of source variables in a manner consistent with respect to the source statement where execution has halted. To facilitate error correction, we have developed a set of engineering change tools; i.e. constraint manipulation algorithms which perform functional or timing modifications on the design, while minimally altering its original optimized specification.; Since business models based on intellectual property are vulnerable to a number of potentially devastating obstructions such as misappropriation and IP fraud, we have developed two techniques for protecting copyright. The first technique, design watermarking, uses constraint manipulation to embed a secret message in the optimized design solution. The second technique, forensic analysis, analyzes whether a solution to an optimization problem could have been created using a particular optimization algorithm.