Social Network Enhanced Device-to-Device Communication Underlaying Cellular Networks

Device-to-device (D2D) communication has seen as a major technology to overcome the imminent wireless capacity crunch and to enable new application services. In this paper, we propose a social-aware approach for optimizing D2D communication by exploiting two layers: the social network and the physical wireless layers. First we formulate the physical layer D2D network according to users’ encounter histories. Subsequently, we propose an approach, based on the so-called Indian Buffet Process, so as to model the distribution of contents in users’ online social networks. Given the social relations collected by the Evolved Node B (eNB), we jointly optimize the traffic offloading process in D2D communication. In addition, we give the Chernoff bound and approximated cumulative distribution function (CDF) of the offloaded traffic. In the simulation, we proved the effectiveness of the bound and CDF. The numerical results based on real traces show that the proposed approach offload the traffic of eNB’s successfully.

💡 Research Summary

This paper, titled “Social Network Enhanced Device-to-Device Communication Underlaying Cellular Networks,” presents a novel framework for optimizing D2D communication by leveraging the interplay between social networks and physical wireless networks. The core objective is to alleviate the traffic burden on cellular base stations (eNBs) by intelligently offloading data traffic to direct D2D links between user equipments (UEs), guided by social behavior patterns.

The proposed framework operates on two layers: the Offline Social Network (OffSN) and the Online Social Network (OnSN). The OffSN models the physical D2D communication network. It is constructed based on historical user encounter data (contact duration and frequency). A probabilistic “closeness metric” is calculated for each pair of UEs, representing the likelihood of establishing a successful D2D session, derived from a fitted Gamma distribution of contact durations. UEs with a closeness metric exceeding a dynamic threshold are clustered into an OffSN, ensuring a reliable physical substrate for D2D communication among socially-connected users who frequently meet.

The OnSN models the virtual space where content dissemination occurs, such as on social media platforms. To predict content popularity and user selection behavior, the authors employ the Indian Buffet Process (IBP), a Bayesian nonparametric model. In this analogy, users are customers, online contents are dishes, and the probability of a user selecting a content is influenced by a base preference and the number of previous users who have selected it. This allows the eNB to maintain and update a dynamic probability distribution over contents, identifying which ones are “old” (already cached within the OffSN) and which are “new.”

The traffic offloading algorithm integrates these two models. When a user requests content, the eNB first checks the user’s location. If the user is in a “white” area (not part of any OffSN), the eNB serves the request directly. If the user is within an OffSN, the eNB consults the OnSN model. If the requested content is predicted to be “old,” the eNB identifies the content holder within the same OffSN who has the highest closeness metric to the requester and orchestrates a D2D communication link between them. If the D2D link fails, the eNB reverts to serving the user. For “new” content requests, the eNB serves directly, and the content becomes a candidate for future D2D offloading.

For performance evaluation, the authors provide a theoretical analysis of the offloaded traffic. They derive the Chernoff bound for the number of offloaded content requests, offering a probabilistic upper limit on performance. Furthermore, they approximate the probability mass function (PMF) and cumulative distribution function (CDF) of this offloaded traffic, providing a statistical characterization of the algorithm’s output. Simulations using real-world mobility traces demonstrate the effectiveness of the proposed approach, showing significant traffic reduction at the eNB. The numerical results also validate that the derived Chernoff bound and approximated CDF accurately reflect the simulated performance, confirming the robustness and predictability of the social-aware D2D offloading scheme.