Secure dynamic adaptive traffic masking

Abstract

As public internetworks are increasingly used for secure communications, the need grows for end-to-end protection fi'om traffic analysis. The additional protection of Traffic Flow Confidentiality can be detrimental to performance when padding is used to mask traffic patterns. Traffic masking policies that are responsive to system service requirements can improve performance, but secure adaptive traffic masking has to balance performance requirements with system protection requirements. This paper addresses the information leaks that result J~om adaptations in security mechanisms. Introduction and BackgroundTraffic flow confidentiality (TFC) is concemed with hiding communication patterns that, if exploited, could reveal or compromise sensitive information. Sources of traffic flow information that need to be protected are frequency and length of transmittals, origin/destination traffic patterns, and protocol headers [6,7]. TFC is becoming more important as government agencies and private companies are moving away from private networks and using open data networks to meet their needs. While the security of open data networks is a concern, designers of network security are faced with an explosion of worldwide communications that includes increased data rates, universal connectivity, new services, and higher standards for performance. In such environments TFC can meet the growing need for protection from traffic analysis, but can be expensive because traffic masking involves the use of padding. Secure dynamic adaptive traffic masking (S-DATM) contributes to a global vision, providing the capability of operating in a commercial environment via traffic protected by appropriate levels of TFC with minimal impact on other traffic.

Timing Analysis in Low-Latency Mix Networks: Attacks and Defenses

“…Timmerman [29,30] proposed an adaptive "traffic masking" technique for latency-insensitive applications such as email. Her S-DATM algorithm uses cover traffic and artificial delays to ensure that traffic emerging from each user conforms to a certain profile.…”

Survey on anonymous communications in computer networks

“…A robust anonymous communication system should be able to safeguard traffic analysis. The research in this area includes traffic flow confidentiality [103][104][105] and traffic flow analysis [106][107][108][109].…”

Formal methods for cryptographic protocol analysis: emerging issues and trends

“…An anonymizing protocol depends upon a mix of traffic to disguise the source and destination of any particular item. Statistical studies, such as the work of Timmerman [82] or Wright et al [85], will be of use here to determine how well this strategy works in different situations. Previous work done on the analysis of covert channels in networks [84] might also be of use here, if applied properly.…”