Optimal Geocast Scheduling under Multicasts and Relaying in mmWave Vehicular Networks

Abstract

Due to the increasing volume of data generated by sensors in modern cars, the need for high data rate links rises in vehicular networking. A promising way to achieve this is using mmWave communications, although beamforming is needed to overcome the high propagation losses at these frequencies. In addition, relaying might be needed to extend the coverage to larger distances for the delivery of a message in a specific geographical area, called geocasting. Moreover, reaching multiple receivers at once via multicasting can be achieved by using a wider antenna beamwidth, which comes at the cost of transmission range, while spatial sharing can be exploited using narrow beams. This paper investigates if using multicasts is beneficial for routing and scheduling of mmWave geocasts that need to be delivered before a timeout. We consider a non-time-slotted system with realistic antenna model and multiple data rates, for which we seek an optimal solution by modeling it as a mixed-integer linear program. Our numerical evaluations show that using multicasts is especially advantageous in scenarios with multiple highway lanes. Furthermore, we devise a heuristic algorithm that efficiently finds a route and creates a transmission schedule for a geocast. Several methods to include multicast links are evaluated, of which some consistently outperform the unicast-only method.