Efficient rate adaptation algorithm in high-dense vehicular ad hoc network

Abstract

In the Vehicular Ad hoc NETwork (VANET) communications, the presence of intersections, vehicles, and obstacles such as buildings may obstruct signal propagation especially in urban areas. Moreover, the vehicle's mobility and density also cause intermittent in inter-vehicle connection due to unpredictable and diverse channel conditions. Rate adaptation is the key method in VANET to predict the prevailing channel conditions and decide appropriate transmission rate quickly based on channel status. Although various rate adaptation algorithms using closed and open loop exist, it is difficult to select the best rate adaptation method due to unpredictable nature of the vehicular environment, while inappropriate data selection leads poor network performance as well. This paper aims to evaluate the performance of different data rate adaptation algorithms in VANET, using closed-loop and open-loop based approaches. The two-targeted open loops are Atsushi Onoe Algorithm and Minstrel, while Collision-Aware Rate Adaptation (CARA) and Robust Rate Adaptation Algorithm (RRAA) are closed-loop based approaches. The mentioned algorithms are simulated under high-density VANET scenario with a number of nodes starts from 40 up to a hundred using the discrete event network simulator NS-3. Moreover, the distance between the nodes is investigated also over the selected approaches. Simulation results, in general, show that the close-looped outperform the open-looped approach, with RRAA better than others by at least 9.4% and nearby reach to 62% in the average throughput over the different node density as the distance varied. The packet delivery ratio is slightly diverse over the node density. However, an end-end delay is 233.39 ms which is the worst among others, but still adequate for the high critical demand applications.