SORT: A Self-ORganizing Trust Model for Peer-to-Peer Systems

SORT: A Self - Organizing Trust Model

for Peer-to-Peer Systems

Abstract:

Open nature of peer-to-peer systems exposes them to malicious activity. Building trust relationships among peers can mitigate attacks of malicious peers. This paper presents distributed algorithms that enable a peer to reason about trustworthiness of other peers based on past interactions and recommendations. Peers create their own trust network in their proximity by using local information available and do not try to learn global trust information. Two contexts of trust, service, and recommendation contexts, are defined to measure trustworthiness in providing services and giving recommendations. Interactions and recommendations are evaluated based on importance, recentness, and peer satisfaction parameters. Additionally, recommender’s trustworthiness and confidence about a recommendation are considered while evaluating recommendations. Simulation experiments on a file sharing application show that the proposed model can mitigate attacks on 16 different malicious behavior models. In the experiments, good peers were able to form trust relationships in their proximity and isolate malicious peers.

Introduction:

PEER-TO-PEER (P2P) systems rely on collaboration of peers to accomplish tasks. Ease of performing malicious activity is a threat for security of P2P systems. Creating long-term trust relationships among peers can provide a more secure environment by reducing risk and uncertainty in future P2P interactions. However, establishing trust in an unknown entity is difficult in such a malicious environment.

Furthermore, trust is a social concept and hard to measure with numerical values. Metrics are needed to represent trust in computational models. Classifying peers as either trustworthy or untrustworthy is not sufficient in most cases. Metrics should have precision so peers can be ranked according to trustworthiness. Interactions and feedbacks of peers provide information to measure trust among peers. Interactions with a peer provide certain information about the peer but feedbacks might contain deceptive information. This makes assessment of trustworthiness

a challenge.

Existing System:

In the presence of an authority, a central server is a preferred way to store and manage trust information, e.g., eBay. The central server securely stores trust information and defines trust metrics. Since there is no central server in most P2P systems, peers organize themselves to store and manage trust information about each other. Management of trust information is dependent to the structure of P2P network. In distributed hash table (DHT)-based approaches, each peer becomes a trust holder by storing feedbacks about other peers .Global trust

Aim of the Project:

We propose a Self-ORganizing Trust model (SORT) that aims to decrease malicious activity in a P2P system by establishing trust relations among peers in their proximity. No a priori information or a trusted peer is used to leverage trust establishment. Peers do not try to collect trust information from all peers. Each peer develops its own local view of trust about the peers interacted in the past. In this way, good peers form dynamic trust groups in  their proximity and can isolate malicious peers. Since peers generally tend to interact with a small set of peers, forming trust relations in proximity of peers helps to mitigate attacks in a P2P system.

Proposed System:

We implemented a P2P file sharing simulation tool and conducted experiments to understand impact of SORT in mitigating attacks. Parameters related to peer capabilities (bandwidth, number of shared files), peer behavior (online/ offline periods, waiting time for sessions), and resource distribution (file sizes, popularity of files) are approximated to several empirical results. This enabled us to make more realistic observations on evolution of trus

relationships. We studied 16 types of malicious peer behaviors, which perform both service and recommendation- based attacks. SORT mitigated service-based attacks in all cases. Recommendation-based attacks were contained except when malicious peers are in large numbers, e.g., 50 percent of all peers. Experiments on SORT show that good peers can defend themselves against malicious peers without having global trust information. SORT’s trust

metrics let a peer assess trustworthiness of other peers based on local information. Service and recommendation contexts enable better measurement of trustworthiness in providing services and giving recommendations.

Requirement Analysis:

Software Requirements

       Language               :  Java1.5

       Front End               :  Java Swing

       Back End               :  Sql Server 2000

      Operating System   :  Windows Xp.

Hardware Requirements

         Hard disk                   :         60GB

         RAM                :        1GB

                        Processor                  :         P IV

CONCLUSION

A trust model for P2P networks is presented, in which a peer can develop a trust network in its proximity. A peer can isolate malicious peers around itself as it develops trust relationships with good peers. Two context of trust, service and recommendation contexts, are defined to measure capabilities of peers in providing services and giving recommendations. Interactions and recommendations are considered with satisfaction, weight, and fading effect parameters. A recommendation contains the recommender’s own experience, information from its acquaintances, and level of confidence in the recommendation.

These parameters provided us a better assessment of trustworthiness. Individual, collaborative, and pseudonym changing attackers are studied in the experiments. Damage of collaboration and pseudospoofing is dependent to attack behavior. Although recommendations are important in hypocritical and oscillatory attackers, pseudospoofers, and collaborators, they are less useful in naive and discriminatory attackers. SORT mitigated both service and recommendation-based attacks in most experiments. However, in extremely malicious environments such as a 50 percent malicious network, collaborators can continue to disseminate large amount of misleading recommendations. Another issue about SORT is maintaining trust all over the network. If a peer changes its point of attachment to the network, it might lose a part of its trust network. These issues might be studied as a future work to extend the trust model