No abstract

-In this paper we present IDAMN, a distributed system whose main functionality is to track and detect mobile intruders in real-time. IDAMN includes two algorithms which model the behaviour of users in terms of both telephony activity and migration pattern. The main novelty of our architecture is its ability to perform intrusion detection in the visited location and within the duration of a typical call as opposed to existing designs that require the reporting of all call data to the home location in order to perform the actual detection. The algorithms and the components of IDAMN have been designed in order to minimize the overhead incurred in the fixed part of the cellular network.

No abstract

The mcreasing development of mobile netwOT'ks r'aises new security requirements and concerns. In addition to the basic need of authentication, confidfntiality and key distribution services, a new problem involving privacy is the unauthor'ized tracking of user's' migration. In other words, accessing any information r'elated to the mobile user's location data without his consent, is a serious violation of his privacy. Mor'eover, if no care is taken, the disclosure of the mobile user real identity may appear during the authentiwtion process. The basic solution to this problem is the use of aliases which insure non-traceability by hiding the user's real identity and also his relationship with domain authorities. In this paper we pr'ovide a classification of the differ'ent degr'ces of non-tmcfllbility and present a new efficient method for' thf computation of aliases. This technique can bc u8cd during authndiClltion of mobile users and thus avoids the drawbacks of existing solutions such as GSM and CDPD.

No abstract