On striking the balance between the fairness of service and throughput in roadside units based vehicular Ad Hoc networks

Data dissemination using Road Side Units (RSUs) in Vehicular Ad Hoc Networks (VANETs) got considerable attention to assist the inter-vehicles communication to overcome the vehicle-to-vehicle frequent disconnection problem. An RSU invokes the underlying scheduler to choose a data item to broadcast in order to satisfy the pending requests of vehicles. Conventionally, a scheduler selects a data item based on different metrics, which are usually the deadline of a request, the size of an item and the popularity of an item. In the conventional scheduling manner, normally the popular data items are broadcast many times for maximising the broadcast channel bandwidth. As a consequence, non-popular data items are broadcast very few times, which creates unfairness to the pending requests for the non-popular data items. However, the non-popular data items may also be important for a vehicle. Hence, there is trade-off of achieving higher fairness and gaining higher system throughput. In this study, we investigate this trade-off in the context of RSUbased VANETs and propose a fairness-friendly approach with which the integration of a scheduling algorithm can balance the trade-off of the fairness of service and the system throughput. Simulation results support our proposed approach and offer the expected results.