Title: Machine learning for efficient link adaptation strategy in VANETs

Authors: Etienne Alain Feukeu; Lukas W. Snyman

Addresses: Department of Electrical Engineering, College of Science, Engineering and Technology (CSET), University of South Africa, Pretoria, Gauteng, South Africa ' Institute for Nanotechnology and Water Sustainability, College of Science, Engineering and Technology (CSET), University of South Africa, Pretoria, Gauteng, South Africa

Abstract: The benefit brought by Vehicular Ad Hoc Networks (VANETs) can only be gained if the successful Road State Information (RSI) message notifications are exchanged between the mobiles involved. Moreover, a successful exchange is only possible with a well-integrated Link Adaptation (LA) mechanism. Furthermore, the higher mobility induces Doppler Shift (DS) in the carrier frequency component of the transmitter node, which corrupts the transmitted signal and makes decoding difficult at the receiver end. Several authors have addressed the LA in VANETs, but almost all of them have done so without incorporating an effective DS mitigation strategy. The current study presents a Machine Learning (ML) approach for an efficient LA strategy in VANETs. The simulation results demonstrated that the ML outperformed AMC, ARF and Cte in threefold, with an improvement level of 212% in terms of throughput, 86.5% in terms of transmission duration and 39% in terms of model efficiency.

Keywords: VANET; Doppler shift; machine learning; link adaptation; WAVE; DSRC; V2V; V2I; IEEE802.11p.

DOI: 10.1504/IJVICS.2023.132930

International Journal of Vehicle Information and Communication Systems, 2023 Vol.8 No.3, pp.279 - 307

Received: 09 May 2022
Received in revised form: 10 Aug 2022
Accepted: 22 Aug 2022
Published online: 17 Aug 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article