Abstract
Distributed systems, such as structured P2P overlay networks, are vulnerable to malicious peers. This is due to their flexibility, which allows to any given single participant to appear as many. This will lead, unfortunately, to the possibility of allowing such participant to control much more peers than it can actually do. This is commonly known as Sybil attack. On the other hand, the voluntary arrival and departure of thousands or millions of peers creates a collective effect called Churn. Consequently, an attacker could exploit this way by generating peers which joining and leaving the network very quickly in order to corrupt the best function of the network. Moreover, an attacker could generate a pollution attack leading to corrupt or delete data stored in the system. In such situation, these attacks enable the malicious participant to mediate and monitor most overlay traffics, enabling arbitrarily the denial of service or censorship attacks. Unfortunately, most of existing security solutions are incapable or ineffective to detect malicious peers in dynamic and decentralized systems such as structured P2P systems. Hence, we propose in this paper a Monitoring Framework, which attempts to secure structured P2P networks by monitoring, analyzing, detecting, and mitigating misbehavior peers.