基于深度特征学习的网络流量异常检测方法

针对网络流量异常检测过程中提取的流量特征准确性低、鲁棒性差导致流量攻击检测率低、误报率高等问题,该文结合堆叠降噪自编码器(SDA)和softmax,提出一种基于深度特征学习的网络流量异常检测方法。首先基于粒子群优化算法设计SDA结构两阶段寻优算法:根据流量检测准确率依次对隐藏层层数及每层节点数进行寻优,确定搜索空间中的最优SDA结构,从而提高SDA提取特征的准确性。然后采用小批量梯度下降算法对优化的SDA进行训练,通过最小化含噪数据重构向量与原始输入向量间的差异,提取具有较强鲁棒性的流量特征。最后基于提取的流量特征对softmax进行训练构建异常检测分类器,从而实现对流量攻击的高性能检测。实验结果表明:该文所提方法可根据实验数据及其分类任务动态调整SDA结构,提取的流量特征具有更高的准确性和鲁棒性,流量攻击检测率高、误报率低。     

Mahalanobis Distance Based Approach for Anomaly Detection of Analog Filters Using Frequency Features and Parzen Window Density Estimation

Analog filters play a very important role in insuring the availability of electronic systems. Early detection of anomalies of analog filters can prevent the impending failures and enhance reliability. The complex architecture and the tolerances of multiple components make it very difficult to detect anomalies in analog filters. To address this concern, A Mahalanobis distance (MD) based anomaly detection method for analog filters is proposed in this paper. The conventional frequency features and the moment of frequency response are selected as the feature vector. Mahalanobis distance is used to transform the frequency feature vector to one dimensional MD data. The anomaly detection threshold is obtained based on probability density of the health MD data sets which is estimated by Parzen window density estimation method. The efficiency of the proposed method has been verified by two case studies. In the case studies, a comprehensive indicator constructed by miss alarm and false alarm is used to obtain an optimal anomaly detection threshold. One class SVM (OCSVM) based anomaly detection method is used as a comparison with our approach. The results illustrate that: (1) the proposed frequency features can effectively clarify the degradation of analog filters; (2) the proposed MD based approach can detect anomalies in analog filters effectively at an early time stage. (3) the proposed MD based approach can detect anomalies in analog filters more accurately than OCSVM based method.     

基于混合高斯变分自编码网络的异常检测算法

异常数据是指偏离大量正常数据点的数据,往往会对各类系统产生负面影响,存在较大风险。异常检测作为一种有效的防护手段,能够检测数据中的异常,为各类系统的正常运转提供重要支撑,具有重要的现实意义。提出了一种基于混合高斯变分自编码网络的异常检测算法,该算法首先使用混合高斯先验构建变分自编码器,对输入数据进行特征提取,然后以混合高斯变分自编码器构建深度支持向量网络,压缩特征空间,并寻找最小超球体分离正常数据和异常数据,通过计算数据特征到超球体中心的欧氏距离衡量数据的异常分数,并以此进行异常检测。最后在基准数据集MNIST和Fashion-MNIST上评估了该算法,平均AUC分别达到了0.954和0.937。实验结果表明,所提出的算法取得了较好的异常检测效果。     

数字图像模糊滤波盲取证算法

数字图像模糊滤波操作常用于美化润饰“伪造”图像.针对常用的均值滤波、空域高斯低通滤波与中值滤波,提出了一种能同时检测上述3种操作的盲取证算法.首先将高频残差作为特征提取域,然后分别基于二值局部模式LBP和自回归模型提取特征,最后使用支持向量机构造模糊滤波检测器.实验结果表明,所提算法能有效地检测模糊滤波操作,对抗JPEG压缩的鲁棒性能较好.     

基于序贯蒙特卡罗滤波的MIMO快时变信道跟踪

高数据传输速率以及终端的高速移动,导致无线通信信道具有时间选择性与频率选择性两个特征.本文主要研究了基于训练序列的多输入多输出（MIMO）时变频率选择性衰落信道的估计与跟踪问题.首先,根据时变无线信道的动态性,将信道冲击响应近似看作一个低阶的自回归矢量过程（AR）,以便于进行时变信道的跟踪.接着在此模型的基础上,利用序贯蒙特卡罗滤波对MIMO通信系统中的双选择性信道进行了跟踪;跟踪过程中需要与信号检测交替进行,即在状态变量的预测和新息修正的中间要进行一次码元的检测,所采用的方法是极大似然序列检测,最后与扩展卡尔曼滤波作了比较.仿真结果表明,在信道噪声是非高斯的情形下,序贯蒙特卡罗滤波的跟踪性能更优越于扩展卡尔曼滤波.     

被动传感器组网模糊综合贴近度的数据关联算法

针对被动传感器采样非周期且采样数据缺乏距离信息等特点,提出了一种用于解决目标航迹与传感器量测相关联的模糊综合贴近度的数据关联算法.由于被动传感器的量测没有距离信息且传感器探测范围小,本算法首先设置两个关联波门进行量测筛选;然后采用航向确定法得出航向角信息,并综合方位角、俯仰角信息,使用模糊综合的方法进行最终的关联,以解决关联错误率高的问题;最后使用扩展卡尔曼滤波进行目标状态与协方差的更新.实验结果证明了该算法方法的有效性.     

A fuzzy-logic based optical sensor for online weld defect-detection

This paper describes an intelligent optical sensor for real time defect detection in gas metal arc welding processes. The sensor measures the radiations emitted by the plasma surrounding the welding arc, and analyzes the information in real time to determine an index of local quality of the weld. The data processing algorithm encompasses a Kalman filter to reduce the heavy amount of noise affecting the measured signals, and an intelligent fuzzy system to assess the degree of acceptability of the weld. The fuzzy system is also able to detect the risk of specific problems (e.g., anomalies in the current, voltage or speed of the arc, contamination with other materials, holes) and the position of defects along the welding line. In an extensive experimental comparison, the fuzzy system outperforms a former version of the detection algorithm based on a statistical approach.     

一种基于高斯粒子滤波的RAIM算法研究

针对在强非线性条件下扩展卡尔曼滤波 RAIM 算法(EKF-RAIM)性能下降的问题， 本文提出了一种基于高斯粒子滤波的 RAIM 算法(GPF-RAIM)。GPF-RAIM 采用高斯粒子进行非线性状态近似估计，在递推时按高斯分布重新生成新的粒子集合，能够解决粒子的退化问题， 不需要进行重采样步骤，保持了粒子的多样性。仿真结果表明，GPF-RAIM 能够有效的检测伪距跳变，相较于 EKF-RAIM 方法，可以获得更小的状态估计误差，提高检测性能。     

Joint angular and spectral estimation technique using nonlinear Kalman filters for cognitive radio

The joint estimation of direction of arrivals (DOA) and carrier frequencies of band-limited source signals is considered in this paper. A novel technique based on nonlinear Kalman filters is proposed for this joint angular and spectral estimation problem for cognitive radio (CR). Since sampling a wideband spectrum at Nyquist rate increases the analog-to-digital converter (ADC) requirements, we propose executing Kalman filter algorithm over a spatial state space model. Thus, one time sample is required and hardware complexity is reduced. Two types of nonlinear Kalman filters, extended Kalman filter (EKF) and unscented Kalman filter (UKF), are proposed. We consider their sub-optimal performance and show how to control their convergence. However, the proposed algorithms can detect a number of source signals limited to the number of elements in employing arrays.     

Clustering‐based resource discovery on Internet‐of‐Things

Resource discovery on Internet‐of‐Things paradigm is an eminent challenge due to data‐specific activities with respect to foraging and sense‐making loops. The prerequisite to deal with the challenge is to process and analyze the data that require resources to be indexed, ranked, and stored in an efficient manner. A novel clustering technique is proposed to resolve the specified challenge. The technique, namely, iterative k‐means clustering algorithm, targets concrete cluster formation using similarity coefficients of vector space model and performs efficient search against matching criteria with respect to complexity. It is simulated on MATLAB, and the obtained results are compared with fuzzy k‐means and fuzzy c‐means clustering algorithm with similarity coefficients of vector space model against exponential increase in the number of resources.     

惯性导航系统/测距机系统辅助的机载全球导航卫星系统欺骗式干扰自主检测算法

全球导航卫星系统(GNSS)欺骗导致目标接收机生成错误的定位结果。利用惯性导航系统(INS)辅助,基于卡尔曼滤波新息序列构造卡方检验统计量是检测机载GNSS欺骗的有效手段。然而,该算法无法给出欺骗的持续时间,从而导致INS/GNSS系统无法依据该算法判断其解算的定位信息是否正确。该文结合测距机系统(DME),提出一种基于重构新息序列的有限记忆卡方检测算法。该算法使用已有的INS,GNSS和DME数据构造一种不参与卡尔曼滤波的新息序列,然后将该新息序列构造成有限记忆卡方检验统计量,从而实现对欺骗式干扰的检测。仿真表明,当机载GNSS欺骗造成250 m及以上的位置偏差时,所设计的算法能够获得较为准确的欺骗持续时间。最后,该文依据所提算法的检测结果,给出了INS/GNSS/DME系统正确的定位信息。     

Adaptive Kalman filter equaliser

A new adaptive equaliser for digital communication channels is proposed. It is derived from a state-variable representation of the communication channel by means of the extended Kalman filter technique. The equaliser proposed is structured as two Kalman filters, one estimating the transmitted data symbols, the other the unknown channel parameters.     

基于深度学习的网络异常检测和智能流量预测方法

边缘计算场景下,边缘设备时刻产生海量蜂窝流量数据,在异常检测任务中针对直接对原始数据检测异常存在的计算冗余问题,提出基于特征降维的蜂窝流量数据异常检测方法.该方法在全局范围内利用LSTM自编码器提取流量数据特征和标识异常网格,然后在存在可疑异常的网格使用K?means聚类进行局部异常确认,结果表明可以更好地检测出不同活...     

基于鲁棒卡尔曼滤波的盲自适应多用户检测算法

提出一种新的基于虚拟噪声补偿技术的鲁棒卡尔曼滤波估计异步CDMA系统多用户接收器的最优判决向量的方法,构造出一种收敛速度快、跟踪性能好、数值稳定性好的高性能盲自适应多用户检测算法。仿真实验表明, 该文提出的方法具有很强的抗多址干扰能力和较高的数值鲁棒性。     

基于边缘特征的SAR图像目标检测方法

如何快速而准确地检测出SAR图像中的目标是一个极富挑战性的课题.利用图像边缘特征和模糊集理论设计了一种快速有效的SAR图像目标检测算法.该算法先利用模糊软阈值小波降噪方法去除相干斑噪声,然后用模糊边缘检测器检测出降噪图像的边缘,最后利用形态学操作算子提取出边缘图中的目标区域.与基于亮度特征以及基于纹理特征的检测算法相比,提出的检测算法能够快速、准确地检测出目标,而且产生的虚警数量较少.SAR实测数据的实验结果表明,提出的算法是有效的且具有很好的应用前景.     

一种混合模型的时序数据异常检测方法

对设备性能指标、用户数据指标的异常检测能有效地发现系统潜在故障,本文提出了一种混合异常检测方法。该方法利用k-means将历史数据按照时间进行划分,使用grubbs算法剔除历史数据中的噪音,并计算各时间段的阈值形成动态阈值,同时利用曲线拟合和ARIMA算法对预处理后的历史数据进行训练,得到对应的模型,作为判断异常的依据。该方法结合了统计学的高效、机器学习的准确,无需对数据进行标注,该方法能自动发现单指标和多指标异常。通过在几个系统的实际运维的检验,本文提出的方法能有效地发现缺数异常和系统异常,提高告警准确率,单指标的查全率达到100%,平均查准率为95.7%,算法的效率满足生产环境中的性能要求。     

基于大数据的电力信息网络流量异常检测机制

随着智能电网建设的加强,电力信息网络及其承载的业务系统得到迅猛发展,网络业务流量的检测和预警具有重要的安全意义.针对目前电力信息网络缺乏处理流量异常问题的有效技术手段,提出了一种基于大数据的电力信息网络流量异常检测机制,并通过对改进的局部异常因子(M-LOF)和支持向量域数据描述(SVDD)两种常用异常检测算法的对比分析,总结出适合电力信息网络的流量异常检测方法.     

GAD:topology-aware time series anomaly detection

To solve the problems of anomaly detection,intelligent operation,root cause analysis of node equipment in the network,a graph-based gated convolutional codec anomaly detection model was proposed for time series data such as link delay,network throughput,and device memory usage.Considering the real-time requirements of network scenarios and the impact of network topology connections on time series data,the time dimension features of time series were extracted in parallel based on gated convolution and the spatial dependencies were mined through graph convolution.After the encoder composed of the spatio-temporal feature extraction module encoded the original input time series data,the decoder composed of the convolution module was used to reconstruct the time series data.The residuals between the original data and the reconstructed data were further used to calculate the anomaly score and detect anomalies.Experiments on public data and simulation platforms show that the proposed model has higher recognition accuracy than the current time series anomaly detection benchmark algorithm.     

基于GSM网络的老年人身体姿态检测系统设计

论文设计了一套基于GSM(Global system for mobile communication,GSM)网络的老年人身体姿态检测系统,该系统以Cortex-M3系列STM32芯片为控制器模块、MMA7361加速度传感器为姿态检测模块、ATK-NEO-6M GPS为位置定位模块、GTM900C为信息发送模块,并运用卡尔曼滤波和支持向量机算法检测老年人跌倒行为。实验结果表明,该系统能准确判断人体正常活动与跌倒事件,能自动定位并发送短消息。     

基于小波和扩展卡尔曼的心电检测

文中将小波变换和扩展卡尔曼滤波器相结合,利用小波变换多尺度多分辨的特点,将心电信号进行分解。然后对心电信号在各尺度上进行扩展卡尔曼滤波。最后在扩展卡尔曼滤波的输出结果上进行QRS波形检测。文中方法经MIT-BIH心电数据库检验,QRS波Se（探测灵敏度）在99.40%以上,同时,QRS＋P（正探测率）在99.39%以上,提高了心电信号检测的正确率。