摘要
Due to their unique characteristics, such as the dynamic changing topology, the absence of central management, the cooperative routing mechanisms, and the resources constraints, Mobile ad hoc networks (MANETs) are relatively vulnerable to both active and passive attacks. In MANET, routing attacks try to disrupt the functions of routing protocol by intentionally or unintentionally dropping packets or propagating faked routing messages. However, due to their computation requirements, the prevention mechanisms are not powerful enough to secure MANET. In this paper, we propose a distributed and cooperative scheme using statistical methods to detect routing attacks in MANETs. Our scheme uses both direct and indirect observations to characterize the behaviors of both neighboring and remote nodes. Simple threshold and Grubb’s Test are utilized to propose our new detection methods. The scheme includes innovative methods to compute our proposed measures, Maximum Accusation Number (MAN) and Accusation Number (AN), which are used to make decision about node’s behavior. Experimental results show that our scheme performs well in detecting anomalous events in routing functions.
Due to their unique characteristics, such as the dynamic changing topology, the absence of central management, the cooperative routing mechanisms, and the resources constraints, Mobile ad hoc networks (MANETs) are relatively vulnerable to both active and passive attacks. In MANET, routing attacks try to disrupt the functions of routing protocol by intentionally or unintentionally dropping packets or propagating faked routing messages. However, due to their computation requirements, the prevention mechanisms are not powerful enough to secure MANET. In this paper, we propose a distributed and cooperative scheme using statistical methods to detect routing attacks in MANETs. Our scheme uses both direct and indirect observations to characterize the behaviors of both neighboring and remote nodes. Simple threshold and Grubb’s Test are utilized to propose our new detection methods. The scheme includes innovative methods to compute our proposed measures, Maximum Accusation Number (MAN) and Accusation Number (AN), which are used to make decision about node’s behavior. Experimental results show that our scheme performs well in detecting anomalous events in routing functions.
