Tehnički vjesnik, Vol. 23 No. 5, 2016.
Izvorni znanstveni članak
https://doi.org/10.17559/TV-20150207070245
Subjective logic based trust model for geographic routing in mobile ad hoc networks
A. Rajesh
; S. K. P. Engineering College, Department of Computer Science and Engineering, Thiruvannamalai – 606 611, Tamil Nadu, India
V. Raji
; S. K. P. Engineering College, Department of Computer Science and Engineering, Thiruvannamalai – 606 611, Tamil Nadu, India
N. Mohan Kumar
; S. K. P. Engineering College, Department of ECE, Thiruvannamalai – 606 611, Tamil Nadu, India
Sažetak
In the past decades, several trust models have been proposed to enhance the security of Mobile Ad hoc Networks (MANETs). The conventional trust models fail to express the notion of uncertainty during the establishment of trust relationships between mobile nodes. Furthermore, they lack in considering the context attributes in trust evaluation. This paper proposes a subjective logic based trust model in POR (SLT-POR) that integrates both the behavioural and context based trust in trust model. The behavioural based trust incorporates subjective logic based evidence fusion in indirect trust evaluation to explicitly represent and manage ignorance as uncertainty. The trust relationship between nodes cannot always reflect the actual relationship and consequently the executed decision from the extracted trust relationship is not always accurate. The subjective logic theory includes a dynamic base rate operator that evaluates an expectation of an opinion and handles the uncertainty in evidence collection. In context, attributes based trust measurement, the SLT-POR estimates a context based trust value using the contextual attributes of each node in the trusted positive progress set. The SLT-POR considers the battery power to include node’s lifetime in the trustworthiness estimation. It assigns a weight for both behaviour and context based trust to efficiently calculate total trustworthiness of a node. The simulation results demonstrate that the packet delivery ratio of the proposed SLT-POR is high even if 50 % of total nodes are malicious and outperforms Context-Aware Security and Trust (CAST) framework.
Ključne riječi
behavioural-based trust; contextual attributes; MANETs; POR; subjective logic; trust model
Hrčak ID:
167494
URI
Datum izdavanja:
13.10.2016.
Posjeta: 1.459 *