Cyber-Physical System (CPS) is a new kind of digital technology that increases its attention across academia, government, and industry sectors and covers a wide range of applications like agriculture, energy, medical, transportation, etc. The traditional power systems with physical equipment as a core element are more integrated with information and communication technology, which evolves into the Cyber-Physical Power System (CPPS). The CPPS consists of a physical system tightly integrated with cyber systems (control, computing, and communication functions) and allows the two-way flows of electricity and information for enabling smart grid technologies. Even though the digital technologies monitoring and controlling the electric power grid more efficiently and reliably, the power grid is vulnerable to cybersecurity risk and involves the complex interdependency between cyber and physical systems. Analyzing and resolving the problems in CPPS needs the modelling methods and systematic investigation of a complex interaction between cyber and physical systems. The conventional way of modelling, simulation, and analysis involves the separation of physical domain and cyber domain, which is not suitable for the modern CPPS. Therefore, an integrated framework needed to analyze the practical scenario of the unification of physical and cyber systems. A comprehensive review of different modelling, simulation, and analysis methods and different types of cyber-attacks, cybersecurity measures for modern CPPS is explored in this paper. A review of different types of cyber-attack detection and mitigation control schemes for the practical power system is presented in this paper. The status of the research in CPPS around the world and a new path for recommendations and research directions for the researchers working in the CPPS are finally presented.

/pdf/cyber-physical-power-system-cpps-a-review-on-modeling-378wefi3hb.pdf

Cyber-Physical Power System (CPPS): A Review on Modeling, Simulation, and Analysis With Cyber Security Applications

With the rapid development of network technology, active defending of the network intrusion is more important than before. In order to improve the intelligence and accuracy of network intrusion detection and reduce false alarms, a new deep neural network (NDNN) model based intrusion detection method is designed. A NDNN with four hidden layers is modelled to capture and classify the intrusion features of the KDD99 and NSL-KDD training data. Experiments on KDD99 and NSL-KDD dataset shows that the NDNN-based method improves the performance of the intrusion detection system (IDS) and the accuracy rate can be obtained as high as 99.9%, which is higher when compared with other dozens of intrusion detection methods. This NDNN model can be applied in IDS to make the system more secure.

https://ietresearch.onlinelibrary.wiley.com/doi/epdf/10.1049/iet-ifs.2018.5258

Network intrusion detection algorithm based on deep neural network

A cyber-physical system (CPS) typically consists of the plant, sensors, actuators, the controller and a communication network. The communication network connects the individual components to achieve the computing and communication in the CPS. It also makes the CPS vulnerable to network attacks. How to deal with the network attacks in CPSs has become a research hotspot. This paper surveys the types of network attacks in CPSs, the intrusion detection methods and the attack defense strategies. The future research directions of CPSs network security are also presented.

/pdf/a-survey-of-network-attacks-on-cyber-physical-systems-3zx11nzez4.pdf

A Survey of Network Attacks on Cyber-Physical Systems

An imbalanced class problem occurs within abundant real-world applications, e.g., fraud detection, text classification, and cancer diagnosis. Beside balancing the imbalanced data distribution to deal with imbalanced data problems, another significant way to solve the bias-to-majority problem is via proper feature selection. This work is intended to use a feature selection method that can choose a subset of features and make ROC AUC and F-measure results in order to achieve high performance on a minority class. In this paper, a weighted Gini index(WGI) feature selection method is proposed. In order to evaluate the proposed method, a comparison result among Chi-square, F-statistic and Gini index feature selection is shown, and Xgboost is the classifier that is used to test the performance of the subset of features. Experimental results indicate that F-statistic contains the best performance when a few features are selected. However, when the number of selected features increases, WGI feature selection achieves the best results. A comparison between the average results from ROC AUC and F-measure are also presented. It shows that ROC AUC always contains a good performance, even if only a few features are selected, and only changes slightly as the subset of features expands. However, the performance of F-measure achieves a good performance after 60% of features are chosen. The results are helpful for practitioners to select a proper feature selection method when facing a practical problem.

Weighted Gini index feature selection method for imbalanced data

In this paper, we review and analyze intrusion detection systems for Agriculture 4.0 cyber security. Specifically, we present cyber security threats and evaluation metrics used in the performance evaluation of an intrusion detection system for Agriculture 4.0. Then, we evaluate intrusion detection systems according to emerging technologies, including, Cloud computing, Fog/Edge computing, Network virtualization, Autonomous tractors, Drones, Internet of Things, Industrial agriculture, and Smart Grids. Based on the machine learning technique used, we provide a comprehensive classification of intrusion detection systems in each emerging technology. Furthermore, we present public datasets, and the implementation frameworks applied in the performance evaluation of intrusion detection systems for Agriculture 4.0. Finally, we outline challenges and future research directions in cyber security intrusion detection for Agriculture 4.0.

Cyber Security Intrusion Detection for Agriculture 4.0: Machine Learning-Based Solutions, Datasets, and Future Directions

Wireless networks are more vulnerable to cyberattacks than cable networks. Compared with the misuse intrusion detection techniques based on pattern matching, the techniques based on model checking U+0028 MC U+0029 have a series of comparative advantages. However, the temporal logics employed in the existing latter techniques cannot express conveniently the complex attacks with synchronization phenomenon. To address this problem, we formalize a novel temporal logic language called attack signature description language U+0028 ASDL U+0029. On the basis of it, we put forward an ASDL model checking algorithm. Furthermore, we use ASDL programs, which can be considered as temporal logic formulas, to describe attack signatures, and employ other ASDL programs to create an audit log. As a result, the ASDL model checking algorithm can be presented for automatically verifying whether or not the latter programs satisfy the formulas, that is, whether or not the audit log coincides with the attack signatures. Thus, an intrusion detection algorithm based on ASDL is obtained. The case studies and simulations show that the new method can find coordinated chop-chop attacks.

An intrusion detection algorithm for wireless networks based on ASDL

Linear Temporal Logic (LTL) Model Checking (MC) has been applied to many fields. However, the state explosion problem and the exponentially computational complexity restrict the further applications of LTL model checking. A lot of approaches have been presented to address these problems. And they work well. However, the essential issue has not been resolved due to the limitation of inherent complexity of the problem. As a result, the running time of LTL model checking algorithms will be inacceptable if a LTL formula is too long. To this end, this study tries to seek an acceptable approximate solution for LTL model checking by introducing the Machine Learning (ML) technique. And a method for predicting LTL model checking results is proposed, using the several ML algorithms including Boosted Tree (BT), Random Forest (RF), Decision tree (DT) or Logistic Regression (LR), respectively. First, for a number of Kripke structures and LTL formulas, a data set A containing model checking results is obtained, using one of the existing LTL model checking algorithm. Second, the LTL model checking problem can be induced to a binary classification problem of machine learning. In other words, some records in A form a training set for the given machine learning algorithm, where formulas and kripke structures are the two features, and model checking results are the one label. On the basis of it, a ML model M is obtained to predict the results of LTL model checking. As a result, an approximate LTL model checking technique occurs. The experiments show that the new method has the similar max accuracy with the state of the art algorithm in the classical LTL model checking technique, while the average efficiency of the former method is at most 6.3 million times higher than that of the latter algorithms, if the length of each of LTL formulas equals to 500. These results indicate that the new method can quickly and accurately determine LTL model checking result for a given Kripke structure and a given long LTL formula, since the new method avoids the famous state explosion problem.

/pdf/ltl-model-checking-based-on-binary-classification-of-machine-7ybwe3s33j.pdf

LTL Model Checking Based on Binary Classification of Machine Learning

Metallic surface defect detection is critical to ensure the quality of industrial products. Recently, human-advanced surface defect detection algorithms have been proposed. Most of these algorithms rely on convolutional neural networks (CNN) and an anchoring scheme. However, a convolution unit only samples the input feature maps at fixed shapes and locations. Similarly, a set of anchors are uniformly predefined with fixed scales and shapes, which increases the difficulties of bounding box regression. Therefore, we propose an adaptive convolution and anchor network for metallic surface defect detection, named ACA-Net. Specifically, an adaptive convolution and anchor (ACA) module is proposed, which mainly consists of adaptive convolution and an adaptive anchor. Firstly, an adaptive convolution module (ACM) is designed, which adaptively determines the location and shape of each convolution unit. In addition, a multi-scale feature adaptive fusion (MFAF) is proposed, which is used in ACM to extract and integrate multi-scale features. Then, an adaptive anchor module (AAM) is proposed to yield more suitable anchor boxes by adaptively adjusting shapes. Extensive experiments on NEU-DET dataset and GC10 dataset validate the performance of the proposed approach. ACA-Net achieves 1.8% on NEU-DET dataset higher Average Precision (AP) than GA-RetinaNet. Furthermore, the proposed ACA module is also adopted in GA-Faster R-CNN, improving the AP by 1.2% on NEU-DET dataset.

/pdf/aca-net-an-adaptive-convolution-and-anchor-network-for-2rtt5in3.pdf

ACA-Net: An Adaptive Convolution and Anchor Network for Metallic Surface Defect Detection

Object tracking is aimed at tracking a given target that is only specified in the first frame. Due to the rapid movement and the interference of cluttered backgrounds, object tracking is a significant challenging issue in computer vision. This research put forward an innovative feature pyramid and optical flow estimation based on the Siamese network for object tracking, which is called SiamFP. The SiamFP jointly trains the optical flow and the tracking task under the Siamese network framework. We employ the optical flow network based on the pyramid correlation mapping to evaluate the movement information of the target in two contiguous frames, to increase the accuracy of the feature representation. Simultaneously, we adopt spatial attention as well as channel attention to effectively restrain the ambient noise, stress the target area, and better extract the features of the given object, so that the tracking algorithm has a higher success rate. The proposed SiamFP obtains state-of-the-art performance on OTB50, OTB2015, and VOT2016 benchmarks while exhibiting better real-time and robustness.

/pdf/real-time-object-tracking-algorithm-based-on-siamese-network-tgv52tw2.pdf

Miaolei Deng

Papers

An intrusion detection algorithm for wireless networks based on ASDL

LTL Model Checking Based on Binary Classification of Machine Learning

ACA-Net: An Adaptive Convolution and Anchor Network for Metallic Surface Defect Detection

Real-Time Object Tracking Algorithm Based on Siamese Network

Random Search and Code Similarity-Based Automatic Program Repair