Title: Research on dynamic trust model based on recommendation trust iteration

Authors: Fengyin Li; Shuang Zhao; Jian Zhou

Addresses: School of Information Science and Engineering, Qufu Normal University, Rizhao, Shandong, 276826, China ' School of Information Science and Engineering, Qufu Normal University, Rizhao, Shandong, 276826, China ' School of Information Science and Engineering, Qufu Normal University, Rizhao, Shandong, 276826, China

Abstract: For the difficulty to accurately calculate the recommendation trust value of nodes in current P2P trust management models, it is hard to recognise the malicious recommendations of some nodes. By introducing the concept of historical successful transaction rate and the concept of the best recommendation chain into trust calculation process, a dynamic trust management model based on recommendation trust iterations is proposed in this paper. In the new trust model, the recommendation trust value of nodes can be more accurately calculated. Therefore the trust management model could describe the dynamic trusts of nodes, and it could help users to make decisions to select the transaction nodes, against malicious nodes and malicious recommendations. Simulation results show that the novel trust model has better dynamic adaptability and higher security compared with the current models.

Keywords: trust management; trust iteration; recommendation trust; best recommendation chain; dynamic trust models; P2P; peer-to-peer; malicious recommendations; malicious nodes; simulation; adaptability; security.

DOI: 10.1504/IJCSM.2016.076037

International Journal of Computing Science and Mathematics, 2016 Vol.7 No.1, pp.20 - 28

Received: 22 Jul 2015
Accepted: 22 Oct 2015

Published online: 22 Apr 2016 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article