Detecting Truthfulness of Packet Dropping Attacks using Public Auditing System in Wireless Ad-Hoc Networks

Multi-hop wireless ad-hoc network gives increased coverage and provide several benefits over traditional wireless local area networks. This architecture makes it more vulnerable to internal attacks from compromised nodes. One of them is packet dropping attack which is a crucial issue in networks. Link error and malicious packet dropping are two sources for packet losses. While observing a sequence of packet losses in the network, it is difficult to identify whether the loss is due to link errors or malicious nodes. This paper focuses on the insider-attack case, whereby malicious nodes that are part of the route selectively drop a small amount of packets which are critical to the network performance. The malicious node may identify the importance of various packets and then it drops few packets which are important to the network operation. Since packet dropping rate in this case is comparable to the channel error rate, existing detection algorithms cannot achieve satisfactory detection accuracy in identifying packet loss rate. Detection accuracy can be improved by exploiting the correlations between lost packets. In this paper, a public auditing system is used which allows the detector to verify the truthfulness of the packet loss information. The proposed mechanism is privacy preserving, collusion proof, and it incurs low communication and storage overheads at intermediate nodes. The proposed mechanism achieves better detection accuracy than the conventional methods such as a maximum-likelihood based detection.