Trust-Oriented Partner Selection in D2D Cooperative Communications

To meet the rapid growing requirement of data services and offload the heavy base station load, device-to-device (D2D) communication is proposed as one of the key technologies for future cellular network, which can enable users to communicate directly and thus effectively improve the bearing capacity of the network. However, when D2D users transmit packets in the manner of cooperative communication, the cooperative users with non-cooperative behavior, such as the selfishness of user, may lead to a sharp decline in network performance. For this reason, in our paper, a trust-oriented partner selection mechanism (TPSM) is proposed to avoid choosing those users with non-cooperative behavior. In particular, the related psychological researches show that the psychology structure of users can be divided into three aspects: cognition, emotion, and behavior. Considering the psychology structure of users, we build multi-dimensional trust relationships between sending users and cooperative users by evaluating the cognition trust, emotion trust, and behavior trust. Furthermore, aiming at the multi-dimensional trust relationships between sending users and cooperative users, we classify the cooperative users into three categories, namely reliable users, observed users, and unreliable users, by combining the decision-theoretic rough sets based on Naive Bayes. Finally, based on different cooperative transmission scenarios, an optimal partner selection mechanism is proposed in this paper. The optimal partner can be chosen from the reliable cooperative users, according to the different scenarios and factors. The numerical results show that the proposed TPSM can effectively identify the selfish users and notably enhance the packet delivery rate.