Solutions for the Deployment of Communication Roadside Infrastructure for Streaming Delivery in Vehicular Networks

• Published in: Journal of network and systems management 2021-07, Vol.29 (3), Article 32
• Main Authors: Silva, Cristiano M., de Souza, Fernanda S. H., Pitsillides, Andreas, Guidoni, Daniel L.
• Format: Article
• Language: English
• Subjects: Autonomous vehicles; Communications Engineering; Computer Communication Networks; Computer Science; Computer Systems Organization and Communication Networks; Digital media; Heuristic; High speed; Information Systems and Communication Service; Infrastructure; Networks; Operations Research/Decision Theory; Pedestrians; Roads; Roadsides; Streaming media; Tabu search; Traffic signals; Traffic speed; Transportation networks; Transportation systems; Travel time; Vehicles
• ISSN: 1064-7570; 1573-7705
• DOI: 10.1007/s10922-021-09600-0

The future of mobility involves the interconnection of the entities of the transportation system (vehicles, roads, traffic lights, pedestrians) in high speed networks providing real-time information to drivers, entertainment for passengers, and a wide variety of applications and systems dedicated to smart transportation. Furthermore, in a few years, autonomous vehicles are going to massively reach the streets, and their interconnection may drastically improve the urban mobility by reducing the travel time and the number of accidents. In this work, we consider the design and management of the network infrastructure for vehicular communication focusing on streaming delivery. We intend to allow a given share of vehicles driving along the road network permanently playing streams received from the network infrastructure, and our main question is where we must provide coverage for achieving a given share of vehicles receiving the media. As parameters, we consider the download data rate that vehicles receive content from the infrastructure, and data consumption rate inside vehicles. An Integer Linear Program formulation along with a tabu search -based heuristic are presented. We consider as baseline the intuitive deployment strategy of covering the most popular locations of the road network. All strategies are evaluated considering a realistic vehicular mobility trace composed of 75, 515 vehicles. Results indicate that the tabu search  heuristic is able to solve a large instance composed of 75, 515 vehicles requiring less covered area than greedy heuristics. Considering the optimal solution, we investigate the solutions on a reduced subset composed of 100 vehicle trips and, considering this reduced scenario, the tabu search  heuristic is able to find the optimal solution.