Charles Wright

Date of Award


Document Type


Degree Name

Doctor of Philosophy (Ph.D.) in Computer Science


Computer Science

Physical Description

1 online resource (ix, 131 pages)


Encryption is an important tool for protecting data, especially data stored in the cloud. However, standard encryption techniques prevent efficient search. Searchable encryption attempts to solve this issue, protecting the data while still providing search functionality. Retaining the ability to search comes at a cost of security, performance and/or utility.

An important practical aspect of utility is compatibility with legacy systems. Unfortunately, the efficient searchable encryption constructions that are compatible with these systems have been proven vulnerable to attack, even against weaker adversary models.

The goal of this work is to address this security problem inherent with efficient, legacy compatible constructions. First, we present attacks on previous constructions that are compatible with legacy systems, demonstrating their vulnerability. Then we present two new searchable encryption constructions. The first, weakly randomized encryption, provides superior security to prior "easily deployable" constructions, while providing similar ease of deployment and query performance nearly identical to unencrypted databases. The second construction, EDDiES, provides much stronger security at the expense of a slight regression on performance.

These constructions show that it is possible to achieve a better balance of security and performance with the utility constraints that come with deployment in legacy systems.

Persistent Identifier