Security Protocol Verification and Optimization by Epistemic Model Checking

Abstract

When a distributed systems protocol is used in a particular context as part of a solution to a larger problem, additional information may be generated from the context. Such information may be used for optimization of the system, and, in the case of security protocols, may be of use to the adversary for attacks. The project conducted a case study of the application of the epistemic model checker MCK to automatically detect such optimization opportunities, and to verify that the protocol remains secure in the mode of use. The particular protocol studied was Chaum's Dining Cryptographers protocol, a security protocol that allows a single agent to anonymously transmit a signal. The context of use considered was a more general 2-phase protocol for anonymous broadcast by an arbitrary number of agents, also proposed by Chaum. The aims of the 2-phase anonymous broadcast protocol were formulated as a knowledge-based program, and an iterative process of model checking and manual counter-example guided refinement was followed to converge on implementations of this knowledge-based program in which local predicates were identified that correspond precisely to the knowledge conditions in the knowledge-based program. This analysis demonstrated that the 2-phase protocol contains some quite subtle flows of information that can be used to optimize its performance, but no violation of the anonymity property was found. As an additional contribution of the research, a formal abstraction technique was developed, and proved correct, for epistemic model checking of protocols that call the Dining Cryptographers protocol as a subroutine. Experimental results show that the optimization improves epistemic model checking performance by orders of magnitude and enables problems of larger scale to be attacked.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Nov 05, 2010
Accession Number
ADA534019

Entities

People

  • Ron Van Der Meyden

Organizations

  • University of New South Wales

Tags

DTIC Thesaurus Topics

  • Abstracts
  • Air Force
  • Air Force Research Laboratories
  • Case Studies
  • Computer Programming
  • Computer Science
  • Concrete
  • Construction
  • Cybersecurity
  • Language
  • Optimization
  • Programming Languages
  • Security
  • Security Protocols
  • Specifications
  • Standards
  • Verification

Fields of Study

  • Computer science

Readers

  • Applied Combinatorial Optimization and Logic Circuit Design.
  • Computer Networking
  • Mathematical Modeling and Probability Theory.