Publication Date

12-2015

Date of Final Oral Examination (Defense)

10-29-2015

Type of Culminating Activity

Thesis - Boise State University Access Only

Degree Title

Master of Science in Computer Science

Department

Computer Science

Major Advisor

Jyh-haw Yeh, Ph.D.

Advisor

Dianxiang Xu, Ph.D.

Advisor

Tim Anderson, Ph.D.

Abstract

Virtualization of the database is the most sought out feature in recent years. Large companies have started to outsource their data to providers who maintain them. Database as a service or data outsourcing is the new paradigm for database management, which helps the user avoid buying expensive hardware and software.

Economic benefits are the main reason for outsourcing databases, but it has few concerns and issues like data security and data integrity. Database-as-a-service (DaaS) has been commercialized in cloud industry such as Sales-force’s Database.com, Amazon’s Relational Database Service, Heroku’s SQL DaaS, and Google’s Google Cloud SQL, etc. However, outsourcing data makes it vulnerable to security breaches and data attacks. In addition to data privacy, data integrity is another important concern that slows down the adoption of this new emerging DaaS technology. Protecting the data in its original form is called data integrity. Data integrity attacks come in different flavors that involve data manipulation, data addition or data deletion.

The challenge to protect the data integrity in outsourced databases is the main motivation for this thesis research. Integrity Coded Databases (ICDB) is a scheme to convert the regular database to integrity coded database that helps in identifying the correctness of the data and also identifies the old data attack. In this thesis, we have used the RSA signature to generate the integrity codes for the database so that the database owner can verify whether the information is correct and fresh from the queried data. However ICDB’s come with memory and performance issues which are discussed in the comparative analytical results.

Share

COinS