TRADING: Traffic Aware Data Offloading for Big Data Enabled Intelligent Transportation System
Tasneem Darwish, Kamalrulnizam Abu Bakar, Omprakash Kaiwartya, Jaime Lloret
Abstract
Todays' Intelligent Transportation System (ITS) applications majorly depend on either limited neighbouring traffic data or crowd sourced stale traffic data. Enabling big traffic data analytics in ITS environments is a step closer towards utilizing significant traffic patterns and trends for making more precise and intelligent decisions particularly in connected autonomous vehicular environments. Towards this end, this paper presents a Traffic Aware Data Offloading (TRADING) approach for big traffic data centric ITS applications in connected autonomous vehicular environments. Specifically, TRADING balances offloading data traffic among gateways focusing on vehicular traffic and network status in the vicinity of gateways. In addition, TRADING mitigates the effect of gateway advertisement overhead to liberate the transmission channels for traffic big data transmission. The performance of TRADING is comparatively evaluated in a realistic simulation environment by considering gateway access overhead, load distribution among gateways, data offloading delay, and data offloading success ratio. The comparative performance evaluation results show some significant developments towards enabling big traffic data centric ITS.