A SPIN-based model checker for telecommunication protocols

Abstract

Telecommunication protocol standards have in the past and typically still use both an English description of the protocol (sometimes also followed with a behavioural SDL model) and an ASN.1 specification of the data-model, thus likely making the specification incomplete. ASN.1 is an ITU/ISO data definition language which has been developed to describe abstractly the values protocol data units can assume; this is of considerable interest for model checking as subtyping in ASN.1 can be used to constrain/construct the state space of the protocol accurately. However, with current practice, any change to the English description cannot easily be checked for consistency while protocols are being developed. In this work, we have developed a SPIN-based tool called EASN (Enhanced ASN.1) where the behaviour can be formally specified through a language based upon Promela for control structures but with data models from ASN.1. An attempt is also made to use international standards (X/Open std on ASN.1/C++ translation) as available so that the tool can be realised with pluggable components. One major design criterion is to enable incremental computation wherever possible (for example: hash values, consistency between alternate representations of state). We have used EASN to validate a simplified model of RLC (Radio Link Control) in the W-CDMA stack that imports datatypes from its associated ASN.1 model. In this paper, we discuss the motivation and design of the EASN language, the architecture and implementation of the verification tool for EASN and some preliminary performance indicators.

Thanks are due to Nokia Research Center, Helsinki for funding this work under SID project 99033. We thank Ari Ahtiainen and Markku Turunen of NRC for their initial project formulation and some key ideas in the software engineering aspects, Dinesh Shanbhag for helping us understand the ASN.1/C++ standard, and Matti Luukkainen, University of Helsinki, for many suggestions and criticisms.