Bandhana, A., Kroupa, T., & García, S. (2024, May). Trust in Shapley: A Cooperative Quest for Global Trust in P2P Network. In Proceedings of the 23rd International Conference on Autonomous Agents and Multiagent Systems (pp. 132-140).
Abstract
The domain of digital networking has witnessed a lot of advancements, one of which is peer-to-peer (P2P) networks for file sharing. These networks gained significant traction with the success of Napster, BitTorrent, Gnutella, and Kazaa [13]. P2P networks offer enhanced scalability and resilience by eliminating the need to go through a server [24]. While the P2P network has transformed the file-sharing landscape, the main challenge these networks face is the ability to determine their peers’ trustworthiness accurately. Ensuring the reliability and authenticity of files being shared within the vast expanse of P2P networks is important, especially when there is no central authority to enforce security protocols. Unlike traditional client-server architecture, P2P systems distribute responsibilities across all peers, meaning each peer acts as a client and a server [1]. These networks are categorized as pure P2P, hybrid, and super-peer networks based on the degree of decentralization [17]. Pure P2P network maintains a flat architecture where all their peers are equal in functionality. The hybrid P2P system combines the elements of both P2P and client-server models [27]. Striking a balance between the two, super-peer networks work on an intermediate layer [5]. While the P2P networks avoid a single point of failure, they become susceptible to security threats and malicious attacks [18]. The network also attracts adversarial peers who may deliberately disrupt the network operations. To keep P2P networks secure, peers use their trust in other peers to protect themselves, block others, or tell others whom to block. Trust is a peer’s belief in another peer’s capabilities, honesty, and reliability based on their own direct experiences [22]. Trust models are pivotal in building any security solution because trust is an inherent concept of any security decision. Until now, most trust models are “I trust you or not” or better,“I trust you until I don’t trust you and is too late”. These models propose a varying degree of trust that changes over time to improve security decisions. A trust model defines the method and procedure of trust modeling and evaluation by assigning a score to each peer [21]. The score is called a local trust if it is based on two peers’ direct interaction, and global trust when it is a peer’s overall standing in the P2P network [9]. There is a wide range of P2P trust models studied in the literature; see [3, 4, 12, 15, 19, 20, 23, 25]. While diverse in their methodologies, these trust models share a common objective: to control potential risks in peer interactions by establishing a robust trust framework. In particular, EigenTrust [11] stands out as one of the prominent methods in this domain.
Read more: https://dl.acm.org/doi/abs/10.5555/3635637.3662860