THIS PAPER HAS BEEN ACCEPTED FOR PUBLICATION IN IEEE INTERNET OF THINGS JOURNAL 1 Context-aware Information Diffusion for Alerting Messages in 5G Mobile Social Networks Giuseppe Araniti 1 , Antonino Orsino 1 , Leonardo Militano 1 , Li Wang 2 , Antonio Iera 1 1 ARTS Lab., DIIES Dept., University Mediterranea of Reggio Calabria, Italy e-mail:[araniti|antonino.orsino|leonardo.militano|antonio.iera]@unirc.it 2 School of Electronic Engineering, Beijing University of Posts and Telecommunications, Beijing, P.R. China e-mail: liwang@bupt.edu.cn Abstract—In emerging Fifth Generation (5G) systems, mobile social networks are expected to play an important role to enable proximity-based content distribution among devices. In this paper, we address social-awareness aspects and device-to- device (D2D) communications for information diffusion solutions in emergency scenarios. Context-aware information is collected from a set of devices deployed in the environment and received data are integrated and elaborated at the cellular base station before being delivered. In such a framework, we model the expected information diffusion time by taking into account both networking- and sociality-related metrics. In particular, we introduce the so-called social inter-contact time which is able to model the interaction frequency between the user and a generic social platform. The proposed approach is compared with alternative solutions where the dissemination process is either managed through direct links from the central base station, as a conventional multicast scheme, or with the support from proximity communications, as a D2D-enhanced multicast scheme. The results of a performance assessment study show that the proposed framework achieves considerable gains, up to 50%, in terms of overall information diffusion time and data rate per UE. Index Terms—Information Diffusion, LTE-A, 5G, D2D, Mul- ticast, Mobile Social Networks, Emergency. I. I NTRODUCTION C URRENT evolutions in wireless networks are slowly changing the way we perceive and interact with the world around us. Future 5G systems are expected to introduce a wide range of innovations that will lead to a “connected reality” where each person will be constantly tethered not only to friends, colleagues, peers and personal services, but also to their vehicles, appliances, stores, home and, in short, to every object that may be of interest [1]. In particular, the fast growth of ”new objects” for sensing actuation and communi- cations, as well as new types of ”mobile” network nodes (e.g., drones), lead the scientific community to specify scenarios that are expected to target the future enabling technology. A remarkable example is given by the scenarios described in [2]: (i) Amazingly fast, (ii) Great service in a crowd, (iii) Best experience follow you, (iv) Super real-time and reliable connections, and (v) Ubiquitous things communicating. Also emergency scenarios, where rescue teams or even common people need to be promptly informed about the surroundings and the security procedures to face a sudden threat, may greatly benefit from the technological enhancements that 5G systems will introduce. Indeed, public-safety and emergency communication networks deeply changed with the adoption of Long Term Evolution (LTE) radio technology and the availability of new multimedia applications and services. At the same time, in several recent tragic events we have wit- nessed how social networks and social media may play a fundamental role in information gathering and in sending out alerting messages. This means that end-users and devices present in a specific disaster area may become source of crucial information in support of context reconstruction and timely rescue operations. Context-awareness can further benefit from Internet of Things (IoT) objects available on the field not necessarily belonging to any specific end-user [3][4]. Logi- cally, the amount of collected information is useless if not elaborated and suitably conveyed to the interested devices, such as those held by rescue teams or single end-users in life- threatening situations. In such a scenario, we believe that social networks are deemed to prospectively play an important role to support dissemination of information among the ”actors” of the system. This paper proposes the use of social ties [5] and of device- to-device (D2D) communications to guarantee an effective gathering of information and its subsequent distribution in a reduced diffusion time. D2D communications between User Equipments (UEs) in mutual proximity are identified as a promising technology to enhance the performance of cellular systems in terms of coverage extension, improved spectrum utilization, higher throughput, and lower energy consumption [6], [7]. In our vision, D2D communications may be a valuable means to activate high performing direct links both in the uplink direction, when collected data must be delivered to the central base station, and in the downlink, when the resulting composite information has to be transmitted to the interested devices. With reference to the aforementioned objectives and refer- ence scenario, the main contributions of the paper are: • We propose a framework where context-aware informa- tion is collected from objects and devices deployed in a certain area of interest. The data is elaborated at the base station and then effectively distributed to the interested users by trying to minimize the information diffusion time. • We investigate the potential offered by the adoption of D2D communications to reduce the data transfer time both in uplink and downlink directions.