Vehicular Data Offloading by Road-Side Units Using Intelligent Software Defined Network

The evolution of wide variety of applications that are used by vehicular users includes a lot of data hungry applications. This increases the workload on the cellular networks, thereby delivering poor service to the users. We can overcome this problem by sharing this workload with open wireless networks. As such, this improves the Quality of Service provided by cellular networks. RoadSide Units (RSU) are a wireless network which plays a major role in data offloading. Our approach discusses switching the communication network from cellular to RSU whenever there is an opportunity for a vehicle to offload vehicles data. Busy roads/urban traffic consists of several RSUs with many users. In urban environment, the vehicular user needs to choose an RSU from several available RSUs within the vehicle communication proximity. For seamless connectivity, the delay in network communication because of selecting the best RSU and frequent switching of connection between vehicles and RUSs must be minimized. In this paper, we propose a Smart Ranking based Data Offloading (SRDO) algorithm for selecting an RSU and to improve the Quality of Service. In SRDO algorithm, Q-Learning is utilized for RSU selection. This algorithm is modelled in Software Defined Network controller to deal with the problem of choosing the RSU in an intelligent way for data offloading. Abstract The evolution of wide variety of applications that are used by vehicular users includes a lot of data hungry applications. This increases the workload on the cellular networks, thereby delivering poor service to the users. We can overcome this problem by sharing this workload with open wireless networks. As such, this improves the Quality of Service provided by cellular networks. RoadSide Units (RSU) are a wireless network which plays a major role in data offloading. Our approach discusses switching the communication network from cellular to RSU whenever there is an opportunity for a vehicle to offload vehicles data. Busy roads/urban traffic consists of several RSUs with many users. In urban environment, the vehicular user needs to choose an RSU from several available RSUs within the vehicle communication proximity. For seamless connectivity, the delay in network communication because of selecting the best RSU and frequent switching of connection between vehicles and RUSs must be minimized. In this paper, we propose a Smart Ranking based Data Offloading (SRDO) algorithm for selecting an RSU and to improve the Quality of Service. In SRDO algorithm, Q-Learning is utilized for RSU selection. This algorithm is modelled in Software Defined Network controller to deal with the problem of choosing the RSU in an intelligent way for data offloading.

Publication year:2020

Keywords:Road-Side Unit, Software Defined Network, Reinforcement Learning, Mobile Edge Computing, Keywords: Road-Side Unit

BOF-keylabel:yes

Accessibility:Open