An Integrated Framework for Formal Verification
and Distributed Simulation
of Asynchronous Hardware


EPSRC Project No. GR/S11091/01 & GR/S11084/01


Partners


Project Summary

This joint proposal aims to develop a practical integrated framework for the analysis of asynchronous hardware designs via formal verification and distributed simulation. The proposal will address weaknesses of the simulation for asynchronous systems (particularly its failure to deal with non-determinacy and deadlock detection) by exploiting the particular suitability of asynchronous systems for formal verification.

The system will be implemented as an enhancement of an existing Balsa synthesis toolkit, developed by the AMULET group at Manchester, that already supports limited analysis of the designs via simulation. The core of the framework will be the process algebra CSP for maximum flexibility and reusability of the methods. CSP has an underlying mathematical theory and inherent parallelism, features that can be exploited to perform both formal verification and distributed simulation.

Previous approaches have been hampered by the fact that verification has been, by necessity, post-hoc. The CAD systems used for the specification and implementation of circuits have been divorced from the verification tools and have totally non-orthogonal underpinnings. In this proposal, Balsa is derived from CSP. Thus, there is the same theoretical basis in the synthesis tool and the verification tool.

Formal verification of properties such as deadlock and equivalence checking will thus be implemented through the translation of Balsa into machine readable CSP, invoking the model checker FDR and back-translating the diagnostics. Distributed simulation will ensure timing analysis of Balsa designs and performance evaluation of benchmark programs. The project will also investigate the applicability of semi-formal verification approaches, hybrid combinations of formal verification and simulation, in the context of asynchronous hardware. The framework will be evaluated on major asynchronous hardware designs, such as the Amulet3i processor aimed at embedded systems.

Aims and Objectives

This project is funded by EPSRC  for three years starting April 2003.

Links


Publications


Case studies


Some conferences