Analytical Model and Performance Evaluation of Long-Term Evolution for Vehicle Safety Services

Wenfeng Li,Xiaomin Ma,Jun Wu,Kishor S. Trivedi,Xin-Lin Huang,Qingwen Liu

DOI: https://doi.org/10.1109/tvt.2016.2580571

IF: 6.8

2017-01-01

IEEE Transactions on Vehicular Technology

Abstract:In a traffic jam or dense vehicle environment, vehicular ad hoc networks (VANETs) cannot meet the safety requirement due to serious packet collisions. The traditional cellular network solves packet collisions but suffers from long end-to-end delay. Third-Generation Partnership Project (3GPP) Long-Term Evolution (LTE) overcomes both drawbacks; thus, it may be used, instead of VANETs, in some extreme environments. We use Markov models with dynamic scheduling and semipersistent scheduling (SPS) to evaluate how many idle resources of LTE can be provided for safety services and how safety applications impact LTE traditional users. Based on the analysis, we propose to reserve the idle radio resources in LTE for vehicular safety services (LTE-V). Additionally, we propose the weighted-fair-queuing (WFQ) algorithm to schedule beacons for safety services using the LTE reserved resources. Numerical results verify that the proposed mechanism can significantly improve the reliability of safety applications by borrowing limited LTE bandwidth. We also build an NS3 simulation platform to verify the effectiveness of the proposed Markov models. Finally, the reliability of applications, including cooperation collision warning (CCW), slow vehicle indication (SVI), and rear-end collision warning (RCW), using dedicated short-range communication (DSRC) with LTE-V, are evaluated. The simulation results demonstrate that the stringent quality-of-service (QoS) requirement of the aforementioned three applications can be satisfied, even under heavy traffic.