|
We have developed mechanisms to address security problems in wireless communications. Firstly, we examined the threats to the security of the WiMax/802.16 broadband wireless access technology. Threats directed to the physical layer and MAC layer were reviewed in details. The likelihood, impact and risk were evaluated according to a threat assessment methodology proposed by the ETSI. Threats were listed and ranked according to the level of risk they represent. This work was used to prioritize the research directions in WiMax/802.16 security. We investigated on Radio Frequency Fingerprinting (RFF) for building intrusion detection systems for wireless and mobile networks. We developed an anomaly-based intrusion detection approach, which incorporates RFF and a multivariate statistical technique. Our approach is to associate a profile of a wireless device with its corresponding MAC address. The success rate is improved by analyzing multiple chronologically ordered data samples. Simulation results support the feasibility of employing RFF and the statistical technique to successfully detect MAC address spoofing. We addressed the problem of detecting a rogue base station (BS) in WiMax/802.16 wireless access networks. The rogue BS attack represents a major denial-of-service threat against wireless networks. We have developed novel algorithms for detecting inconsistencies of signal strength reports when a system is under attack. These algorithms can be used by an intrusion detection system. For the purpose of anomaly-based intrusion detection in mobile networks, we examined the feasibility of using the mobility patterns of users making use of public transportation. We have presented a novel framework making use of an instance based learning technique for classification purposes. An empirical analysis was conducted in order to assess the impact of key parameters. We concluded that it is feasible to use mobility patterns for anomaly-based intrusion detection. Private communications between peers over non private domains is a challenge. We have developed schemes offering secure communications between mobile nodes using virtual private networking technologies based on IP security and mobile IP. Identity concealment is viewed as a standard identity privacy feature. We have extended the UMTS mutual authentication protocol such that the true identity of a MS cannot be discovered by an eavesdropping attacker.
|
