فصلنامه پدافند الکترونیکی و سایبری
سال نهم شماره 3 (پیاپی 35، پاییز 1400)

Visual tracking of microscopic objects is one of the most important studies of dynamic biological processes and requires automated segmentation and tracking methods. It is often limited to the morphology of objects or human study and lacks the automation and scalability to detect objects, track the path of any object, and examine their topology with the detection of related anomalies. This paper presents a fast scalable agent-oriented method for automatic detection, real-time video tracking, simultaneous tracking of microscopic objects, monitoring object behavior, and their topology based on graph theory applicable to the Internet of Things. It has no mentioned restrictions. Its segmentation method is a combination of temporal and spatial changes of the image to detect moving objects and predict their movement path, and the possibility of detecting individual anomalies of the object (death, moving a stop, collision of objects, a sudden departure from and a sudden entry into processing frame). Provides abrupt onset and onset of anomalies (network splitting, batch changes, batch decomposition, batch spacing, attenuation, and network collapse). The results of experimental experiments to track microscopic objects of sperm and birds in 2D images of 3D video film show that it has 99% sensitivity and 97% accuracy of instantaneous detection of objects with 99% detection accuracy. In monitoring and tracking, correlation and collision of sperm objects have an accuracy of 99.8% and in birds due to environmental noise and error detection in rapid topology changes, birds have an accuracy of 88%.

In the secure two-party computation, two parties wish to jointlycompute a function with of their private inputs,whilerevealing only the output.Yao’s garbled circuit protocol is a classic and solutionto this problem.It is well-known that Yao’s protocol is vulnerable to malicious behavior by its participants. The general approach as cut-and-choose techniques that proposes to solve this vulnerables, but cut-and-choose techniques creates new problems within itself including of selective failure attack and consisitecy inputs.In this paper we present a secure two party computation based onexpanded cut‑and‑choose bilateral oblivious transfer protocol and show that proposed protocol solve selective failure attack and  consisitecy inputs in addition our protocol is significantly more efficient and far simpler than previous works in computation complexity, symmetric encryption operations,bandwidth, and  error probability  by input recovery achive.

GSM cellular standard is still widely used worldwide. In this standard, A5 ciphering algorithms are employed for protecting user data. A5/1 and A5/3 are two variants of A5 ciphering algorithms that are proven to be very powerful. Most known attacks on these ciphering algorithms assume some known plaintext data. In this paper, for the first time, a method of plaintext prediction is proposed for SDCCH logical channel. Four possible downlink SDCCH packets, which are RR, UA, SABM, and UI Fill frames, are considered. The matrices of the occurrence positions and probabilities of these packets are learned by observing the network traffic. Four matrices are considered corresponding to four different types of sessions. Experiments on a real-world network show that we can correctly predict on average 2.94 plaintexts for each session. Moreover, the average position of the first correct plaintext in all predicted plaintexts is equal to 1.24. So, the required time for cipher cracking is around 25% more than the time required by an ideal plaintext prediction system

The use of Global Positioning System (GPS)  in highly automated systems is increasing day by day. Therefore, itssecurity of these systems is getting important more and more. The reliability of the obtained position by GPS is in danger by spoofing attacks. A spoofer transmits replicas of authentic satellite signals to force the victim receiver to misjudge the its position estimate. Numerousresearches have been focused on spoofing detection and mitigation in the GPS receivers. In this paper, mitigation of spoofing attack is suggested by using multi-correlator architecture associated with neural network. Spoofing signal is generated by mixing two signals which are produced by authentic GPS signal and its shifted. The results of the simulations which wasperformed in the software defined receiver, indicate the solution was effective in mitigating the spoofing attack. By studying three scenarios of spoofing, the proposed method was evaluated and the results show that the rate of reduction of deception error is 88.42% by using multi-correlation architecture.

Available botnets currently cover a wide range of Internet shipments. Use the net to access the network from infected computers connected to the Internet, remotely. Using research in this field is done based on the signatures with the result of the discovered results, anomalies, traffic behavior, and existing addresses. This method has not been able to detect a high rate at the moment, which is especially useful when it performs its main behavior, or these are methods that have already been forgotten due to need for memory. It is so great that it is practically impossible to do. The purpose of this study is to propose the construction to perform the identification operation, which is presented in this study with Markov chain and without the use of memory because Markov chain in this study does not require storage memory and does not exist based on behavioral analysis. The proposed method is able to perform useful behaviors using incorrect results of the operation better than the previous solutions, because if it examines the form you need, if such conditions do not exist, it will cause a computational overhead. In this research, various criteria such as medium circuit lines, accuracy and precision under consideration are captured, and in other of these proposed methods, as more possible than other existing methods, it is better if performed.

Internet of things (IoT) constitutes networked devices that can gather and exchange information. Thescarcity of the available spectrum used by a large number of devices in IoT is a challenge. The Cognitive Radiotechnology has emerged as a promising solution to overcome spectrum scarcity in a resource-constrained wireless sensor network. The prioritized spectrum access is the key to maintain the efficiency of CRSN. Modelling the prioritized spectrum access policy is a significant need to analyze a CRSN. However, in CRSN, TCP experiences in severe throughput reduction since it cannot differentiate between the packet loss due to SU’s transmission-blocking and packet loss due to congestion. In this paper, two significant events are investigated that caused secondary user blocking.In additiona Discrete-Time Markov chain (DTMC) is proposed to describe the spectrum usage by both primary and secondary users, which is used to estimate the TCP throughput and end to end delay. The experimental resultsbased on the NS2 confirms the accuracy of the proposed model and show that the throughput and the average end-to-end delay are improved based on the proposed DTMC model comparing with some transport protocol in the cognitive radio networks. The performance results through simulation show that the proposed model achieves up to 20% improvement of the throughput comparing with the classical TFRC and TFRC-CR respectively.

One of the most important security challenges with the advance of technology in cyberspace is phishing attacks. Phishing is a type of cyber-attack that always tries to obtain information such as username, password, bank account information, and the like by forging a website, email address and convincing the user to enter this information. Due to the increasing growth of these attacks and the increasing complexity of the type of attack, current phishing detection systems often cannot adapt to new attacks and have low detection accuracy. Graph-based methods are one of the techniques for identifying malicious domains that use the connections between the domain and IP to identify. In this paper, a graph-based phishing detection system using deep learning is presented. The main steps in the proposed method include extracting IP from the domain, defining the relationship between the domains, determining the weights, and converting the data to a vector by the Node2vec algorithm. Then, using CNN and DENSE deep learning models, the classification and identification operations are performed. The experimental results over three different datasets show that the proposed method provides an accuracy of about 99% in identifying malicious domains, which has an acceptable improvement compared to state of the art in this context.

Today, the need for anomaly-based intrusion detection systems is felt more than ever due to the emergence of new attacks and the increase in Internet speed. The main criterion for determining the validity of an efficient intrusion detection system is the detection of attacks with high accuracy. In addition to inability of existing systems to manage growing attacks, also they have high rates of positive and negative misdiagnosis. This paper uses the ID3 decision tree features for anomaly-based intrusion detection systems. Two feature selection methods are also used to reduce the amount of used data for the detection and categorization. The KDD Cup99 dataset was used to evaluate the proposed algorithm. The test results show a detection accuracy of 99.89% for the DoS attack and an average accuracy of 94.65% for all attacks using the decision tree, indicating better values ​​than previous tasks.

The use of social networks is growing increasingly and people spend a lot of their time using thesenetworks. Celebrities and companies have used these networks to connect with their fans and customers andnews agencies use these networks to publish news. In line with the growing popularity of online socialnetworks, security risks and threats are also increasing, and malicious activities and attacks such asphishing, creating fake accounts and spam on these networks have increased significantly. In a fake accountattack, malicious users introduce themselves instead of other people by creating a fake account and in thisway, they abuse the reputation of individuals or companies. This paper presents a new method for detectingfake accounts in social networks based on machine learning algorithms. The proposed method for machinetraining uses Various similarity features such as Cosine similarity, Jaccard similarity, friendship networksimilarity, and centrality measures. All these features are extracted from the graph adjacency matrix of thesocial network. Then, principal component analysis was used in order to reduce the data dimensions andsolve the problem of overfitting. The data are then classified using the Kernel Density Estimationclassification and the Self Organization map and the results of the proposed method are evaluated using themeasure of accuracy, sensitivity, and false-positive rate. Examination of the results shows that the proposedmethod detects fake accounts with 99.6% accuracy which is about 5% better than Cao's method. The rate ofmisdiagnosis of fake accounts also improved by 3% compared to the same method.

Let be a simple and undirected graph with vertex set . A set is called a dominating set of if every vertexoutside is adjacent to at least one vertex of . For any integer , a dominating set is called a -adjacencydominating set of if the induced subgraph contains at least one vertex of degree at most . The minimumcardinality of a - adjacency dominating set of is called the - adjacency domination number of that isdenoted by . In this paper, the study of - adjacency domination in graphs is initiated, and exact values andsome bounds on the - adjacency domination number of a given graph are presented. Furthermore, it isshown that there is a polynomial-time algorithm that computes the -adjacency domination number of agiven tree. Moreover, it is proven that the decision problem associated to the -adjacency domination isNP-complete for bipartite graphs.

The main purpose of wireless sensor networks (WSNs) is to monitor, record and announce specific conditions from different locations and different applications to the well node or end user. Wireless sensor networks have many applications such as patient status monitoring, target tracking, forest and rangeland monitoring, battlefield, and so on. In these networks, energy is one of the inherent limitations. Because the energy consumed is supplied by a battery, which has a limited lifespan. Clustering is one of the best ways to save energy due to data aggregation, and selecting the right clusters increases the lifespan of wireless sensor networks. Since clustering is one of the NP-hard problems, metaheuristic algorithms are suitable for this problem. In this paper, an energy-aware and cluster-based routing method for WSNs with a combination of multilayer perceptron (MLP)neural network algorithm and simulated annealing (SA) is presented. In the proposed method, the simulated annealing metaheuristic algorithm is simulated to determine the cluster head (CH) and multilayer perceptron neural networks are used to determine the members of each cluster. After the clustering process, data is sent from the source node to the well by creating appropriate routing tables among the headers. The simulation results of the proposed method show that this method improves the parameters of energy consumption, package delivery rate and throughput.

Today, due to high costs, it is not possible to perform a comprehensive and complete test on all parts of thesoftware. But if the fault-prone parts are identified before the test, the main focus of the test can be placedon these parts, which leads to cost savings. Identifying fault-prone components is the main purpose ofsoftware fault prediction. A predictive model receives software modules along with their features as inputand predicts which ones are prone to fault. Machine learning techniques are commonly used to constructthese models, the performance of which is highly dependent on the training dataset. Training datasetsusually have many software features, some of which are irrelevant or redundant, and the removal of thesefeatures is done using feature selection methods. In this research, a new method for wrapper-based featureselection is proposed that uses memetic algorithm, random forest technique and a new criterion based onfuzzy inference system. The results show that the proposed fuzzy evaluation criterion has a betterperformance than the existing criteria and improves the performance of feature selection. The final purposeof this research is to achieve a robust model for predicting high performance software faults and thecomparison results show that the proposed model has higher performance than other models.

Today, intrusion detection systems are extremely important in securing computers and computer networks. Correlated systems are next to intrusion detection systems by analyzing and combining the alarms received from them, appropriate reports for review and producing security measures. One of the problems face intrusion detection systems is generating a large volume of false alarms, so one of the most important issues in correlated systems is to check the alerts received by the intrusion detection system to distinguish true-positive alarms from false-positive alarms. The main focus of this research is on the applied optimization of classification methods to reduce the cost of organizations and security expert time in alert checking. The proposed Incrimental Intrusion Detetection Model using Correlator (IIDMC) is tested on a valid test dataset and the results show the efficiency of the proposed model and consequently its high accuracy.