Clarifying Trust in Social Internet of Things
The main aim of this project is to allocate specific tasks to the user based on the Trust among the Social Internet of Things .
We provide protection for the trustee, exploring information from characteristics and better using results of historical assignments, offering two different schemes for transitivity of trust, evaluating trust not only with positive factors but also with negative factors, and adapting to dynamic environments. Experiments are carried out on an experimental IoT network. The performance improvement is reflected in decreased abuse rates of task delegations with the trust mutuality model as well as increased success rates and decreased unavailable rates with the trust transitivity model. The proposed methods of interaction also increase the net profits of the users and make network agents quickly adapt to a changing environment. In the proposed trust model, the trustor and the trustee perform the bilateral evaluation of the trustworthiness. Therefore, the vicious trustor cannot easily obtain services, and the malicious trustee cannot easily involve in trustor’s tasks. Trustor and trustee evaluate each other on four different aspects, i.e., success rate, gain, damage, and cost. It can prevent the malicious nodes from promoting only a single aspect’s value. Also, this model is a characteristic-based model. Once a node behaves maliciously in a task, subsequent evaluations are affected of many other types of tasks that contain one or more the same characteristics. Malicious behaviors can be detected effectively.