Wormhole attacks are among the most severe attacks on mobile ad hoc networks (MANETs). They do not involve message injection or message alteration, can be staged by outsider nodes, and cannot be prevented simply by encrypting network traffic. This paper further develops a wormhole attack discovery technique based on frequency-space analysis of periodic routing messages. The Frequency-based Wormhole Attack Discovery (FWAD) method described in this work is local, does not require specialized hardware or node synchronization, and works with routing messages readily available in networks that use proactive routing protocols. The new concept introduced in this paper is the use of an existing network characteristic, jitter, as a tool for improving security. Two jitter waveforms - keyed jitter and partitioned jitter - that enhance wormhole attack detection with FWAD are described in this paper. In keyed jitter each node's jitter value is taken from a stream generated using a key known to other network nodes. Partitioned jitter has a high-frequency carrier sinusoidal component built into it. These forms of jitter allow frequency-based wormhole attack detection to take advantage of property that would otherwise inhibit its effectiveness. This paper also demonstrates that attackers cannot easily avoid being detected with FWAD.
展开▼
