Algorithms and tools for anonymization of the internet traffic

Collecting network traffic traces from deployed networks is one of the basic steps in under-standing communication networks and ensuring adequate quality of service. Traffic traces may be used for network management, traffic engineering, packet classification, and for analyzing user behavior. They may also be used for identifying and tracking network anomalies and for maintaining network security. For privacy and security reasons, monitored traffic traces should be anonymization before they may be shared. The goal of anonymization is to preserve traffic properties while enforcing the privacy policies. Various tools and techniques have been implemented for trace anonymization such as: Crypto-PAn, Anon-tool, FLAIM, ip2anonip, IP-Anonymous, and TCPdprive. These tools use a number of anonymization algorithms such as black-marker, random permutations, truncation, pseudo-anonymization, and prefix-preserving pseudo-anonymization. In this thesis, we present a survey of anonymization tools and algorithms. We propose and implement a new anonymization tool called Anonym, which allows the user to execute multi-level anonymization and to display analysis results. Finally, we evaluate the impact of the anonymization on the anonymized traffic data.