End-to-end Security with Translation

Abstract

End-to-end security involves two entities communicating at a distance over an insecure communication channel while maintaining many of the security properties of private, in-person communication. End-to-end security transforms an insecure channel into a secure one, such that the entities know who the other is, their communication is not interpretable by anyone observing the channel, and an active attacker that changes the content communicated will be detected. For digital communication it is desirable to have end-to-end security properties. However, not all entities that wish to communicate share a common language. This may be humans from different countries, web services that encode data in different formats, or applications that communicate with different protocols. The ideas they wish to share are common to both entities, but the representation of them is different. In such a case, end-to-end security limits the ability to use common translation methods that would allow communication. This paper discusses different translation approaches in the context of end-to-end security.