Time-Varying Social-Aware Resource Allocation for Device-to-Device Communication

Abstract

With the rapid development of wireless mobile communication, local content sharing has become the emerging demand for users who are geographically close. Device-to-device (D2D) communication allows two users to communicate directly with each other. In order to achieve more effective content distribution and content spread by allocating better spectrum resources to users with better social networking and content diffusion capabilities, we propose an optimization scheme for resource allocation of the D2D communication by utilizing the potential social relations that are embedded in the communication devices. The degree of intimacy between users is abstracted from the call records to quantify social-relation strengths. Considering the time-varying property of social relations, the auto-regressive integrated moving average model is applied to map the call records into time sequence to predict users’ social relations. Besides, instead of individual utility or the overall network utility, each user aims to maximize its social-community utility which takes other social related D2D users into consideration. Potential game is utilized to solve the social-community utility maximization problem of resource allocation for the D2D communication due to its outstanding mapping nature and always has the Nash equilibrium. Finally, a social-aware distributed resource allocation algorithm is proposed, and the algorithm achieves convergence and stability. Numerical results show that our proposed scheme increases the overall utility over 30% compared with coalition game scheme, and over 50% compared with random selection scheme without loss of the fairness.