BadZak: An hybrid architecture based on virtual backbone and software defined network for Internet of vehicles

In this paper, we focus on the dissemination and data transportation in vehicular networks while reducing the communication cost in terms of overhead, energy and cellular traffic. We consider hybrid communication environment vehicles-to-vehicles (V2V), and Vehicles-to-Infrastructure (V2I). This environment is possible through the vehicular heterogeneous networks where vehicles are equipped with both cellular and non-cellular communication interfaces. In order to connect vehicles to Internet with a minimum solicitation of cellular infrastructure, we propose a new hybrid and programmable architecture based on Connected Dominating Sets (CDS), and Software Defined Vehicular Networks (SDVN) approaches. On the one hand, the SDVN paradigm is introduced to make the architecture flexible and programmable. On the other hand, the CDS is used to select dominating nodes that will form the the virtual backbone (VB). The VB is composed by a subset of selected parked vehicles where their role is to connect other vehicles to Internet and to reduce the end-to-end delay. Only vehicles in this backbone use the cellular link and act not only as relayed nodes but also as fog devices which are intermediate nodes between cloud server and end-users. In order to create and to manage this VB, we propose a new algorithm named BadZak based on CDS approach. The BadZak is based not only on the connectivity degree of parked vehicles, but also on their geographic position, and the zone of interest (ZI) as strategy to select dominating nodes. The ZI concept depends on the application and needs to include nodes of the VB covering a large area. We implemented two versions of BadZak algorithm one based on glutton and other optimal approaches, named BadZak-1 and BadZak-2 respectively. These algorithms are compared with well-known one name Cheng Greedy algorithm in different simulation scenarios and different parameters: transmission range, network density, connectivity degree, and the point of interest number.