Polymorphic Security for Databases

Databases are among the most common and vital applications. The data found in these structures requires a specialized security mechanism. Large amounts of data make databases susceptible to language attacks because the data vastly exceeds the unicity distance of the information they contain. Furthermore, the database's operation makes it challenging to implement effective changes to the encryption of resident data. Relational database structures are designed to be uniquely suited for security using polymorphism techniques. Carlson et al. posited the principles of polymorphic encryption and random number generators, suggesting that these techniques apply to many forms of encryption and randomness. This paper targets a mutating encryption structure that enables independent field-level protection of databases and dynamically modifies that protection over time in response to access and changes to the constituent data.