Collusion-resistant sybil attack detection in mobile ad hoc networks

Abstract

A Mobile Ad-Hoc Networks (MANETs) is an autonomous collection of mobile devices that communicate over relatively bandwidth contrived wireless links. They are mainly convenient and appropriate for critical situations, including armed forces, law enforcement as well as emergency preparedness class of operations and catastrophic conditions to replace the damaged infrastructure networks. Initially MANETs were commenced as supervised networks normally having ownership by a sole unit called offline authority, like military. Due to the increase in the mobile communication devices, an entirely self-organized and managed MANET may be produced. Because of openness, MANETs are subject to various adversarial attacks. In entirely self-organized MANETs, nodes are normally hesitant to expend their valuable resources forwarding the packets of other nodes and are therefore liable to evince selfish or occasionally spiteful malicious behavior. This selfishness can deprive network throughput and possibly lead to network segregation. Cooperation enforcement schemes have been proposed to thwart the issue of selfishness primarily to make sure selfish nodes bear the punishment of their bad actions. However, Due to the lack of centralized identity management or centralized Trusted Third Parties in MANETs, nodes can create zero-cost identities without any restrictions and could escape from punishment or detection by simply changing identity to clear all its bad history, known as whitewashing. Spiteful malicious nodes can concurrently create and command many virtual identities to launch an attack, called a Sybil attack. In Sybil attack, a large number of logical identities can be created on a single physical device by a selfish malicious node which gives a false impression to the network that it were different benign nodes and uses them to launch a harmonized attack against the network or a node. In the context of reputation-based schemes, a Sybil attacker can disrupt the detection precision by slandering other good nodes, boosting its own reputation or exchanging fake positive recommendations about one of its quarantined identities. In order to defend against Sybil attacks, Position verification or localization of nodes seems most promising. Localizing a node requires cooperation of other nodes. But, nodes may not always behave cooperatively and may collude in unfriendly environments. Collusion attacks in location verification engage multiple opponents conspiring to cheat the verifiers of the system into believing that there is a node at the specified location. Collusion attack normally takes place when two or more malicious nodes harmonized their potencies to save one or more Sybil nodes, launch a harmonized attack or to disrupt the detection precision. For example, some malicious colluding nodes may support and share positive recommendations for the Sybil identities of other spiteful nodes being evaluated, making it almost impractical to spot such identities as being Sybil. A successful collusion attack often works on the principle that nodes shows itself as reliable and trustworthy and cooperate in some type of interactions, usually direct interaction and then deceive the node in witness interaction, i.e. providing false information about other nodes to support colluding group or defame or degrade other benign nodes. This forged information promotes the colluding group and the victims will interact with it and will be betrayed. In this research project we figure out that if the assessor node employs a multi-dimensional trust model, collusion attack can be averted, i.e. Trust in Direct interaction as well as in witness interactions. The motivation for having two types of trust is that we believe trustworthiness has different independent dimensions. For example, a node that is honest in a direct interaction is not certainly trustworthy in a witness interaction. The sole purpose of indirect trust computation is to determine the trustworthiness of a (unfamiliar) node from the set of recommendations to slight the gap between the acquired recommendation and the real trustworthiness of the target node for detecting collusion. The proposed scheme is designed to compute the trustworthiness of every node, examine the activity pattern of nodes, detect, and thwart collusion and Sybil attacks. A novel and robust trust based Sybil attack detection-resistant to collusion approach is proposed to accurately detect whitewashing and Sybil attacks in the presence of malicious collusion. We will show that detecting Sybil identities in the presence of collusion attack while exhibiting one-dimensional trust model cannot accurately detect Sybil identities. Through the help of extensive simulations and experiments, we are able to demonstrate that our proposed solution detects Sybil or whitewashers‘ new identities with good accuracy and reduces the benefits of collusion even in the presence of mobility.