Security models

[WORK IN PROGRESS]

Hardness assumptions, IND-CPA and IND-CPA^D security, and noise flooding.