Queue based Vehicular Ad Hoc Network Prognostic Offloading Approach

Abstract

Vehicular Ad hoc NETworking (VANET) enables a vehicle to connect with other vehicles and the surrounding devices such as Road Side Units (RSUs) and base-stations through a wireless network. There are challenging issues within VANET environment caused by the high demand of Internet access. These issues include an increase in the vehicle traffic and the necessity of dynamic topologies. Nowadays, the high usage of Internet in vehicles is also increasing the load on the cellular network base-stations. To alleviate the load from the base-stations, vehicles should be able to switch the communication between the cellular network and RSUs to offload the data. When a vehicle is not within the RSU signal range, it is still possible for the vehicle to exchange information using Vehicle-to-Vehicle (V2V) communication. The main aim of this paper is to predict the vehicles topology, identify multiple offloading paths and compute the costs of the identified paths. Towards this end, knowledge defined network is utilized. To deal with connection interruptions in V2V, we develop algorithms for predicting an efficient V2V offloading path using queues. These algorithms make it possible to reduce the response time, improve the resource management of the network and helps in efficient service connectivity.