Covertness Analysis of Subliminal Channels in Legitimate Communication

Abstract

Secure transaction over a network is one of the critical application which employs high-level cryptographic methods to secure information exchange. Secure communication merely relies on the strength of a cryptographic method. However early research in 80’s have explored possibilities of various loopholes in a cryptosystem, one of which is the attacks based on subliminal channel. This was first demonstrated by G. Simmons for DSA Algorithm by planting covert or subliminal data in one of the variables in DSA. Subliminal channel user embeds the covert data in traditional cryptosystem and this is termed as backdoor - an unintended design mathematically planted in any cryptographic algorithm. Such attacks on secure network will compromise the most important attribute of network called “privacy”.

This paper explores all the theoretical boundaries involved with subliminal channels and proposes a new metric called “covertness index” which indicates the strength of the trapdoor or backdoor placed in the communication channels. This index assists in assessing or evolving a detection method to analyze the embodied trapdoor in a communication channel. This paper also proposes certain changes to elliptic curve model for generating randomness which is a vital step in creation of cryptographic keys in public-key cryptosystem.