Applications of sketches in network traffic measurement: A survey

Accurate and timely network traffic measurement is essential for network status monitoring, network fault analysis, network intrusion detection, and network security management. With the rapid development of the network, massive network traffic brings severe challenges to network traffic measurement. However, existing measurement methods suffer from many limitations for effectively recording and accurately analyzing big-volume traffic. Recently, sketches, a family of probabilistic data structures that employ hashing technology for summarizing traffic data, have been widely used to solve these problems. However, current literature still lacks a thorough review on sketch-based traffic measurement methods to offer a comprehensive insight on how to apply sketches for fulfilling various traffic measurement tasks. In this paper, we provide a detailed and comprehensive review on the applications of sketches in network traffic measurement. To this end, we classify the network traffic measurement tasks into four categories based on the target of traffic measurement, namely cardinality estimation, flow size estimation, change anomaly detection, and persistent spreader identification. First, we briefly introduce these four types of traffic measurement tasks and discuss the advantages of applying sketches. Then, we propose a series of requirements with regard to the applications of sketches in network traffic measurement. After that, we perform a fine-grained classification for each sketch-based measurement category according to the technologies applied on sketches. During the review, we evaluate the performance, advantages and disadvantages of current sketch-based traffic measurement methods based on the proposed requirements. Through the thorough review, we gain a number of valuable implications that can guide us to choose and design proper traffic measurement methods based on sketches. We also review a number of general sketches that are highly expected in modern network systems to simultaneously perform multiple traffic measurement tasks and discuss their performance based on the proposed requirements. Finally, through our serious review, we summarize a number of open issues and identify several promising research directions.     

基于高斯混合隐马尔科夫模型与人工神经网络的紧急换道行为预测方法

为了有效降低因驾驶员紧急换道行为而诱发的交通事故，提高道路交通事故链阻断效率，提出一种基于高斯混合隐马尔科夫模型（GMM-HMM）和人工神经网络（ANN）的紧急换道行为预测方法。首先利用GMM-HMM对车辆行驶状态以及驾驶行为连续观察序列进行换道意图辨识，采用ANN预测下一时段的驾驶行为，再预测换道过程中的横向加速度变化率，从而判断紧急换道的危险程度。驾驶员在环仿真实验及实车实验结果表明，该方法预测避险成功率达92.83%，实验避险成功率达90.32%。该方法能有效地对紧急换道行为进行提前警告与干预。     

基于LSTM-RF模型的城市道路交通流预测

交通流预测作为信号协调和出行时间预测等任务的基础,成为了交通领域的研究点。对于交通流预测问题,研究人员提出了多种方法,但这些方法大多只使用交通流数据的时域信息进行交通流预测,忽略了空间相关性对于预测目标路段流的影响,导致预测精度不理想。基于组合模型的思想提出了一种称为LSTM-RF的交通流预测模型。在交通流预测过程中,首先使用LSTM模型提取预测目标路段的时序特征,再将其预测值与采集的相邻上下游路段信息同时作为随机森林模型的输入特征,进行交通流时空相关性分析,获得最终的预测结果。并通过贵阳市车牌识别系统采集的城区132条路段的交通流数据进行实验验证。结果表明:该方法在预测精度上优于单一模型,并且预测误差相比单一模型有明显减少。     

Profiling drunk driving recidivists in Denmark

Drunk drivers are a menace to themselves and to other road users, as drunk driving significantly increases the risk of involvement in road accidents and the probability of severe or fatal injuries. Although injuries and fatalities related to road accidents have decreased in recent decades, the prevalence of drunk driving among drivers killed in road accidents has remained stable, at around 25% or more during the past 10 years. Understanding drunk driving, and in particular, recidivism, is essential for designing effective countermeasures, and accordingly, the present study aims at identifying the differences between non-drunk drivers, drunk driving non-recidivists and drunk driving recidivists with respect to their demographic and socio-economic characteristics, road accident involvement and other traffic and non-traffic-related law violations. This study is based on register-data from Statistics Denmark and includes information from 2008 to 2012 for the entire population, aged 18 or older, of Denmark. The results from univariate and multivariate statistical analyses reveal a five year prevalence of 17% for drunk driving recidivism, and a significant relation between recidivism and the drunk drivers’ gender, age, income, education, receipt of an early retirement pension, household type, and residential area. Moreover, recidivists are found to have a higher involvement in alcohol-related road accidents, as well as other traffic and, in particular, non-traffic-related offences. These findings indicate that drunk driving recidivism is more likely to occur among persons who are in situations of socio-economic disadvantage and marginalisation. Thus, to increase their effectiveness, preventive measures aiming to reduce drunk driving should also address issues related to the general life situations of the drunk driving recidivists that contribute to an increased risk of drunk driving recidivism.     

Traffic video‐based intelligent traffic control system for smart cities using modified ant colony optimizer

Road traffic congestion is a serious problem in today's world and it happens because of urbanization and population growth. The traffic reduces the transport efficiency in the city, increases the waiting time and travel time, and also increases the usage of fuel and air pollution. To overcome these issues this papers propose an intelligent traffic control system using the Internet of Vehicles (IoV). The vehicles or nodes present in the IoV can communicate between themselves. This technique helps in determining the traffic intensity and the best route to reach the destination. The area of study used in this paper is Vellore city in Tamilnadu, India. The city map is separated into many segments of equal size and Ant Colony Algorithm (AOC) is applied to the separated maps to find the optimal route to reach the destination. Further, Support Vector Machine (SVM) is used to calculate the traffic density and to model the heavy traffic. The proposed algorithm performs better in finding the optimal route when compared to that of the existing path selection algorithms. From the results, it is evident that the proposed IoV‐based route selection method provides better performance.     

交通事故的自动判案研究

该文针对法律领域民事案件中的“交通事故”类案件进行研究,期望在该“交通事故”数据集上实现自动判案。从“中国裁判文书网”采集14 000条数据文本,并对数据进行人工标注。基于对数据集的分析,分别对数据进行粗粒度和细粒度分类,粗粒度为4类,细粒度为8类。该文使用了三种模型: 基于SVM的模型、基于BI-GRU的模型和基于Attention+BI-GRU的模型。实验结果表明: 在该数据集上,对数据进行粗粒度分类时,基于Attention+BI-GRU的模型F1值为80.26%,基于SVM的模型为77.24%,基于BI-GRU的模型为72.65%。在细粒度分类时,基于BI-GRU的模型F1值为48.59%,基于SVM的模型为38.29%,基于Attention+BI-GRU的模型为40.87%。     

空间语义增强下的城市交通事故数据可视分析

目的 海量城市交通事故数据可能蕴含有交通事故的空间模式，挖掘出交通事故的空间模式有助于开展交通事故的防治工作。目前交通管理部门虽然记录了交通事故发生地的空间位置信息，但没有对事故发生地进行空间语义描述，从而影响对交通事故空间模式的深入分析。因此，提出一种交通事故数据空间语义增强方法，并设计了一套可视分析系统。方法 基于城市兴趣点来增强交通事故数据的空间语义。以事故发生点为中心获取周围城市兴趣点，使用特征向量刻画兴趣点的数量、类别及其与事故发生点的距离，并称此向量为空间语义特征向量。将空间语义特征向量和相应的交通事故关联，以达到增强其空间语义的目的。然后，基于空间语义特征向量，使用自组织映射聚类算法对交通事故进行聚类分析，根据其空间语义特征将交通事故分为若干类别。最后，通过使用地图视图展示事故点数据、聚类视图和平行坐标视图展示聚类分析的结果及其空间语义特征的可视化方法，对交通事故的空间模式进行分析。结果 针对空间语义增强的交通事故数据以及相关分析任务，有效地使用上述数据分析方法与可视化技术，设计并实现了一套多视图关联的可视分析系统，提供了便捷的交互方式辅助用户分析。通过研发人员和交通警察共同对安徽省合肥市2018年的交通事故数据进行分析，将交通事故发生地划分9类并指出每类地点的空间语义特点，进一步分析出了事故高发区域的空间语义特性。结论 本文提出的交通事故数据空间语义增强方法和可视分析方法可以帮助用户揭示交通事故的空间语义模式，有助于深入分析和认识交通事故的成因，能为交通事故防治相关的城市建设工作提供建议。     

基于空间合作关系的基站流量预测模型

针对传统的自回归积分移动平均（ARIMA）模型和长短时记忆（LSTM）单元在基站流量预测中没有利用基站（BS）间合作关系的问题，提出一种利用由用户群体在不同基站下访问产生的基站合作关系的流量预测（TPBC）算法。首先，通过基站之间的合作关系构建基站合作网络，并对此合作网络进行社区划分得到基站社区；然后，通过格兰杰因果关系检验方法寻找与目标基站同一社区且关系最紧密的若干基站，作为目标基站的合作基站；最后，使用LSTM和词嵌入层（Embedding）搭建混合神经网络，并根据目标基站和合作基站的流量信息进行流量预测。实验结果表明，TPBC在基站流量预测上的均方根误差（RMSE）相比ARIMA和LSTM分别减小了29.19%和27.47%。TPBC能有效提高基站流量预测准确率，在流量卸载和绿色节能等领域具有重要意义。     

小尺度交通信号灯的检测与状态识别

交通信号灯识别包括检测和状态识别，在智能交通系统中发挥重要作用。基于YOLOv3算法提出了一种交通信号灯检测与状态识别模型。针对交通信号灯相较于交通场景中其他目标具有尺度小的特性进行了算法的设计：降低骨干网络的下采样倍率以增加小尺度目标的特征描述能力；通过增大特征图的尺度来改进多尺度特征融合；引入广义交并比作为检测任务的损失函数来改进目标边界框的回归效果。同时，根据交通信号灯本身的特性，使用颜色和形状约束的方法对信号灯进行状态识别和类别验证。最后在公开的Bosch交通信号灯数据集上和实际的城区道路进行了实验验证。实验结果表明，所提出的算法能够提升交通信号灯识别的精度和召回率，识别准确率可以达到90%左右。     

A study on the monitoring model development for quality measurement of internet traffic

The recent trend of integration among new network services such as the long-term evolution (LTE) based on internet protocol (IP) needs reputable analyses and prediction information on the internet traffic. The IP along with increased internet traffics due to expanding new service platforms such as smartphones will reflect policies such as network QoS according to new services. The establishment of monitoring methods and analysis plans is thus required for the development of internet traffics that will analyze their status and predict their future. The paper with the speed of Internet traffic model is developed for monitoring the state of the experiment and verified. The problem is that the proposed service Internet service provider (ISP) to resolve the conflict between the occurrences can be considerably Internet traffic and that the state of data may be helpful in understanding. The paper advancement policy to reflect the network traffic volume of Internet services and users irradiation with increased traffic due to the development and management of the analysis was carried out experimental measurements.