基于DSRC的车联网性能分析

发布时间：2019-01-02 16:29

【摘要】：伴随着道路安全逐渐引起大众的高度重视,车联网的安全应用越来越得到普及。但是安全报文较高的及时性和准确性要求,使得现有的很多协议无从满足。而当下比较流行的DSRC协议栈恰好可以解决车联网应用的QoS问题。本文的目的就是为DSRC协议栈提出一个准确且高效的分析模型,重点在DSRC的数据链路层。期望通过该模型来得到影响网络性能的参数,进而可以通过调整这些参数来提高网络性能,最终更好地满足安全应用的需求。由于DSRC协议栈在其MAC子层使用的是EDCA机制进行信道接入控制,而EDCA机制又是在DCF机制的基础上加入了多级队列分级思想的。所以本文首先介绍了一个既有的基于DCF机制的二维Markov模型,并在该Markov的基础上进行了一系列网络性能的分析。以上述Markov模型为基础,本文基于EDCA机制提出了两个Markov模型:退避窗口变化的二维Markov模型和连续时隙转移的一维Markov模型。在这两个Markov模型的基础上,主要通过数学公式的推导来得到理论分析上的节点吞吐量和节点单跳时延。首先可以由两个Markov模型得到节点的平均碰撞概率和发包概率。然后本文将分为两种情况来讨论最终的节点吞吐量和单跳时延,分别为节点饱和发包情况以及非饱和发包情况。对于饱和发包情况,首先根据吞吐量的定义来计算吞吐量,即单位时间内节点发包大小。然后可以由吞吐量反向计算出节点单跳时延;对于非饱和发包情况,首先根据单跳时延特性计算出单跳时延,以此为基础再计算出节点吞吐量。为了验证上述模型的正确性,本文根据仿真来得到EDCA机制下系统的表现情况,通过对比和参数调整,便可证明本文提出的模型的正确性。在仿真过程中,通过对每种参数进行多次的生成网络拓扑和实验,来尽量减少仿真结果的随机性,保证仿真结果的准确性。通过理论模型分析和仿真实验,最终验证了本文提出的模型的准确性。理论分析数据和仿真实验数据基本吻合。本文提出的理论分析模型得出的网络性能主要受EDCA参数集的AIFS、最小退避窗口和最大退避窗口的影响。网络性能在网络即将进入饱和发包状态时达到最优。
[Abstract]:With the increasing attention of the public on road safety, the safety application of vehicle networking is becoming more and more popular. However, due to the high requirement of timeliness and accuracy of security message, many existing protocols can not be satisfied. At present, the popular DSRC protocol stack can solve the QoS problem of vehicle networking applications. The purpose of this paper is to propose an accurate and efficient analysis model for DSRC stack, focusing on the data link layer of DSRC. It is expected that the parameters that affect the network performance can be obtained by this model, and then these parameters can be adjusted to improve the network performance and finally meet the requirements of security applications. Because the DSRC protocol stack uses the EDCA mechanism to control the channel access in its MAC sublayer, and the EDCA mechanism is based on the DCF mechanism, the multilevel queue hierarchy is added. So this paper first introduces an existing two-dimensional Markov model based on DCF mechanism, and analyzes a series of network performance based on the Markov. Based on the above Markov model, this paper proposes two Markov models based on the EDCA mechanism: the 2-D Markov model with Backoff window variation and the one-dimensional Markov model with continuous slot transition. On the basis of these two Markov models, the theoretical analysis of node throughput and node single-hop delay is obtained by the derivation of mathematical formulas. First, the average collision probability and contract probability can be obtained from two Markov models. Then this paper will discuss the final node throughput and single-hop delay in two cases, namely, the node saturated outsourcing and the unsaturated outsourcing. In the case of saturated contract delivery, the throughput is first calculated according to the definition of throughput, that is, the node size per unit time. Then the single-hop delay can be calculated from the throughput, and the single-hop delay can be calculated according to the characteristic of single-hop delay, and then the node throughput can be calculated based on the characteristic of single-hop delay. In order to verify the correctness of the above model, the performance of the system under the EDCA mechanism is obtained by simulation, and the correctness of the proposed model can be proved by comparison and parameter adjustment. In the process of simulation, the network topology and experiment are generated several times to reduce the randomness of simulation results and ensure the accuracy of simulation results. The accuracy of the proposed model is verified by theoretical model analysis and simulation experiments. The theoretical analysis data are in good agreement with the simulation data. The network performance obtained by the theoretical analysis model presented in this paper is mainly affected by the AIFS, minimum Backoff window and the maximum Backoff window of the EDCA parameter set. The network performance is optimal when the network is about to enter the saturation state.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U495

【相似文献】