S3: Smart City/Environment 2

Chair: Josip Musić, University of Split, FESB, Croatia
13 Jul 2016

S3: Smart City/Environment 2

  1. An IoT-aware AAL System for Elderly People
    Luca Mainetti, Luigi Patrono, Andrea Secco and Ilaria Sergi (University of Salento, Italy)

    The rapid aging of the population occurred in recent years has encouraged the development of several solutions aimed to guarantee a healthy and safe lifestyle to the elderly. In this paper, an Ambient Assisted Living (AAL) system has been designed in order to create better living conditions for older people. In this way, people can living independently longer in their own house with an improved quality of life. The proposed system includes several features. On the one hand, it is able to continuously monitor the health status of the elderly through data coming from heterogeneous sources (i.e., environmental sensors and medical devices). On the other hand, it is able to guarantee outdoor and indoor localization aimed to know the real-time position of the elderly both inside and outside their home. A remote reasoning system processes all collected data with the aim of generating appropriate events and alerts. The architecture was validated from a functional point of view through a proof-of-concept.

  2. SDN-Based Security Framework for the IoT in Distributed Grid
    Carlos Gonzalez; Salim Mahamat Charfadine; Olivier Flauzac; Florent Nolot (University of Reims Champagne-Ardenne & CReSTIC/SYSCOM, France)

    Software-defined networking (SDN), is evolving as a new paradigm for the next network generation architecture. The separation of control plane and data plane within SDN, brings the flexibility to manage, configure, secure, and optimize network resources using dynamic software programs. From a security point of view SDN has the ability to collect information from the network devices and allow applications to program the forwarding devices, which unleashed a powerful technology for proactive and smart security policy. These functions enable the integration of security tools that can be used in distributed scenarios, unlike the traditional security solutions based on an static firewall programmed by an administrator like Intrusion Detection and Prevention System (IDS/IPS). This network programmability may be integrated to create a new communication platform for Internet of Things (IoT). In this paper, we present our preliminary study that is focused on the understanding of an effectiveness approach to build a cluster network using SDN. For cluster interaction, we introduce an optimum path routing protocol that manages the routing tasks over Cluster-SDN. Our proposed scheme is a starting point for some experiments providing perspective over SDN deployment in a cluster environment for IoT. By using network virtualization and OpenFlow technologies to generate virtual nodes, we simulate a prototype system over 500 devices controlled by SDN and it represents a cluster. The results show that the network devices are only able to forward the packets by predefined rules on the controller. For this reason, we propose a method to control the IP header at the application-level to overcome this problem using Opflex within SDN architecture.

  3. An embedded access control system for restricted areas in smart buildings
    Hyuri S. Maciel, Isadora Cardoso, David F. Silva and Clovis G.M. do Nascimento (Federal University of Alagoas, Brazil); Heitor S. Ramos (Universidade Federal de Alagoas – UFAL, Brazil); Joel J. P. C. Rodrigues (Instituto de Telecomunicações, University of Beira Interior, Portugal); Andre L.L. Aquino (Universidade Federal de Alagoas, Brazil)

    This paper presents an embedded system for access control in restricted area through the integration of two authentication strategies: facial recognition and RFID TAGs. Once that face recognition, in an embedded system, is susceptible to failures and RFIDs TAGs can be used by other users, the integration of these two solutions aims to increase the robustness of access control system. This system can be easily deployed into enterprise environments, meetings and offices rooms. Additionally the system proposed does not require an Internet access. Through the evaluations we observed an acceptable execution time and an increase of robustness of access control system, even using simple techniques of face recognition.

  4. Mapping aerial images from UAV
    Dunja Gotovac, Sven Gotovac, and Vladan Papic (University of Split, Croatia)

    This paper implements and evaluates experimental procedure for mapping images acquired by unmanned aerial vehicles (UAV’s). Proposed model requires at least 4 ground control points (GCPs). Aerial images captured using non-metric cameras contain real details of the images. Used non-parametric approach requires a set of corresponding control points obtained from a referenced image and based map. The corresponding control points are applied manually. Two different referenced layers for mapping images were used and compared. The performance of proposed model is evaluated with the real flight data obtained by typical UAV.