Research Article
Trust Prediction Model Based on Deep Learning in Social Internet of Things
@INPROCEEDINGS{10.1007/978-3-030-67514-1_44, author={Yuyao Wen and Zhan Xu and Ruxin Zhi and Jinhui Chen}, title={Trust Prediction Model Based on Deep Learning in Social Internet of Things}, proceedings={IoT as a Service. 6th EAI International Conference, IoTaaS 2020, Xi’an, China, November 19--20, 2020, Proceedings}, proceedings_a={IOTAAS}, year={2021}, month={1}, keywords={Social internet of things Trustworthiness management Deep learning}, doi={10.1007/978-3-030-67514-1_44} }- Yuyao Wen
Zhan Xu
Ruxin Zhi
Jinhui Chen
Year: 2021
Trust Prediction Model Based on Deep Learning in Social Internet of Things
IOTAAS
Springer
DOI: 10.1007/978-3-030-67514-1_44
Abstract
The Social Internet of Things (SIoT) is the result of the development of Internet of Things from intelligence to socialization. In the social internet of things, different nodes can automatically establish social relationships through social networks to obtain the services they need. Trust management is very important to such an open environment. This paper proposes an improved trust management model for social internet of things, which is divided into two parts: the improved node-level trust model and server-level trust model. In this paper, we propose an innovative trust model at the SIoT server-level, by introducing the deep learning model to predict the trust value of the new nodes in the social internet of things, to solve the problem that the network delay may affect the trust value evaluation in the actual social internet of things network. The simulation results show that the model based on deep learning prediction can get more successful transaction experience, and it is still effective against the high proportion of malicious nodes. The system performance is significantly better than the model without deep learning.
