Leveraging high-order statistics and classification in frame timing estimation for reliable vehicle-to-vehicle communications
Leveraging high-order statistics and classification in frame timing estimation for reliable vehicle-to-vehicle communications
复制标题
利用帧时序估计中的高阶统计和分类来实现可靠的车辆间通信
DOI：
10.1049/iet-com.2017.1066
复制
发表时间：
2018
影响因子：
1.6
通讯作者：
Zhang Yanling
中科院分区：
文献类型：
--
作者：
Zhen Li;Qin Hao;Song Bin;Ding Rui;Zhang Yanling
在车辆到车辆（V2V）通信中，由于V2V传播信道的高度动态性质，实现可靠的物理层性能是一项具有挑战性的任务。帧定时估计作为依赖于信道统计的最关键的信号处理过程之一，必须适当地增强以应对这一挑战。本文提出了一种基于IEEE 802.11p标准的帧定时估计方法。通过设计四阶相关函数和差分归一化函数，该定时度量不仅具有可扩展的相关长度，而且对多径效应和大载波频偏具有鲁棒性。从假设检验和分类的角度来看，该方法能有效地提高类可分性准则下正确和错误定时指标的区分度，从而显著提高了定时估计性能.在典型V2V信道模型下的仿真结果与理论分析一致，表明该方法能显著降低虚警和漏检概率，并使帧检测阈值的选取变得更加容易。
In vehicle-to-vehicle (V2V) communications, achieving reliable physical layer performance is a challenging task due to the highly dynamic nature of V2V propagation channels. Frame timing estimation, as one of the most critical signal processing procedures that rely on channel statistics, has to be appropriately enhanced to tackle this challenge. This study presents a novel frame timing estimation scheme based on both the available periodical preambles in IEEE 802.11p standard. By designing the fourth-order statistics-based correlation and differential normalisation functions, the proposed timing metric not only is capable of possessing an extensible correlation length, but also achieves the robustness to multipath effect and large carrier frequency offset. From the standpoints of hypothesis testing and classification, the proposed approach can effectively increase the distinction between correct and wrong timing indexes in terms of the class-separability criteria, and consequently has a significantly improved timing estimation performance compared with the existing methods. Simulation results consist with theoretical analysis under the typical V2V channel model, and demonstrate that the proposed method can significantly reduce both the probabilities of false alarm and missed detection, and make the selection of a suitable threshold for frame detection much easier.