Protection against Buffer Overflow Attacks through Runtime Memory Layout Randomization

To date a number of comprehensive techniques have been proposed to defend against buffer over attacks. In spite of continuing research in this area, security vulnerabilities in software continue to be discovered and exploited. This is because the existing protection techniques suffer from one or mor...

Full description

Saved in:
Bibliographic Details
Published in:2014 International Conference on Information Technology pp. 184 - 189
Main Authors: Kumar, K. Shiva, Kisore, N. Raghu
Format: Conference Proceeding
Language:English
Published: IEEE 01.12.2014
Subjects:
Generators
Hardware
Internet
large scale cyber-attack
Layout
Libraries
memory randomization
program stack
Security
Software
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:To date a number of comprehensive techniques have been proposed to defend against buffer over attacks. In spite of continuing research in this area, security vulnerabilities in software continue to be discovered and exploited. This is because the existing protection techniques suffer from one or more of the following problems: high run time overheads (often exceeding 100%), incompatibility with legacy C and C++ code, not sufficiently fine grained randomization of memory layout and the inability to perform randomization at run time rather than compile time or link time or load time. While security through diversity is a promising technique to defend against large scale cyber attacks, existing techniques are susceptible to information leakage and brute-force attacks, in addition to the short comings indicated above. To overcome the above indicated drawbacks, in this paper we propose Function Frame Run time Randomization (FFRR) technique. FFRR offers memory layout randomization at run time and performs randomization at the level of individual variables on the stack.
DOI:10.1109/ICIT.2014.57