TY - GEN
T1 - ARC
T2 - 10th International Conference on Applied Cryptography and Network Security, ACNS 2012
AU - Athanasopoulos, Elias
AU - Kemerlis, Vasileios P.
AU - Polychronakis, Michalis
AU - Markatos, Evangelos P.
PY - 2012
Y1 - 2012
N2 - HTTP Parameter Pollution (HPP) vulnerabilities allow attackers to exploit web applications by manipulating the query parameters of the requested URLs. In this paper, we present Application Request Cache (ARC), a framework for protecting web applications against HPP exploitation. ARC hosts all benign URL schemas, which act as generators of the complete functional set of URLs that compose the application's logic. For each incoming request, ARC exports the URL, extracts the associated schema, and searches for it in the set of already known benign schemas. In case the schema is not found, the request is rejected, and the event is recorded. ARC can be transparently integrated with existing web applications without any modifications to the server and client code. It is implemented in Google's Go language and uses efficient data structures for storing the URL schemas, imposing negligible computational overhead on the web application server. When running on a 4-core Linux server, ARC can process hundreds of thousands of URL requests per second. A typical URL resolution is in the scale of microseconds.
AB - HTTP Parameter Pollution (HPP) vulnerabilities allow attackers to exploit web applications by manipulating the query parameters of the requested URLs. In this paper, we present Application Request Cache (ARC), a framework for protecting web applications against HPP exploitation. ARC hosts all benign URL schemas, which act as generators of the complete functional set of URLs that compose the application's logic. For each incoming request, ARC exports the URL, extracts the associated schema, and searches for it in the set of already known benign schemas. In case the schema is not found, the request is rejected, and the event is recorded. ARC can be transparently integrated with existing web applications without any modifications to the server and client code. It is implemented in Google's Go language and uses efficient data structures for storing the URL schemas, imposing negligible computational overhead on the web application server. When running on a 4-core Linux server, ARC can process hundreds of thousands of URL requests per second. A typical URL resolution is in the scale of microseconds.
KW - HPP
KW - Web Security
UR - https://www.scopus.com/pages/publications/84863443464
U2 - 10.1007/978-3-642-31284-7_24
DO - 10.1007/978-3-642-31284-7_24
M3 - Conference contribution
AN - SCOPUS:84863443464
SN - 9783642312830
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 400
EP - 417
BT - Applied Cryptography and Network Security - 10th International Conference, ACNS 2012, Proceedings
Y2 - 26 June 2012 through 29 June 2012
ER -