Quantitative analysis of a certified e-mail protocol in mobile environments: A probabilistic model checking approach

Abstract

Formal analysis techniques, such as probabilistic model checking, offer an effective mechanism for model-based performance and verification studies of communication systems' behavior that can be abstractly described by a set of rules i.e., a protocol. This article presents an integrated approach for the quantitative analysis of the Certified E-mail Message Delivery (CEMD) protocol that provides security properties to electronic mail services. The proposed scheme employs a probabilistic model checking analysis and provides for the first time insights on the impact of CEMD's error tolerance on computational and transmission cost. It exploits an efficient combination of quantitative analysis and specific computational and communication parameters, i.e., the widely used Texas Instruments TMS320C55x Family operating in an High Speed Downlink Packet Access (HSDPA) mobile environment, where multiple CEMD participants execute parallel sessions with high bit error rates (BERs). Furthermore, it offers a tool-assistant approach for the protocol designers and analysts towards the verification of their products under varying parameters. Finally, this analysis can be also utilized towards reliably addressing cost-related issues of certain communication protocols and deciding on their cost-dependent viability, taking into account limitations that are introduced by hardware specifications of mobile devices and noisy mobile environments.