Models are important in science. Appropriate models have allowed scientist to make progress. We illustrate this using two examples:
- In 1979 Shamir discussed how to avoid misuse of digital signatures within a company. His scheme was implemented by RSA Inc. We analyze the model and its weaknesses. The proposal of an alternative model to this problem has been the cornerstone for what is now called threshold cryptography, which allows parties to co-sign messages.
- Perfect secrecy is likely the oldest model in information security. It was used to prove the security of the one-time pad. A consequence of the model is that modern ciphertexts are (indistinguishable from) uniform. This implies that encrypted text is immediately visible to a network sniffer. Information hiding addresses that problem. We look back at models used in modern information hiding to conclude that these are inadequate. We propose a new model for perfect information hiding which is inspired by Shannon’s 50 year old model. We analyze this model in details and propose information hiding schemes.
Part of this presentation is based on joint work with Blackburn, Burmester, De Santis, Di Crescenzo, Frankel, Jajodia, Le, Yung, Wild.