فصلنامه پدافند الکترونیکی و سایبری
سال نهم شماره 2 (پیاپی 34، تابستان 1400)

In military telecommunication systems, advanced techniques are used to intercept and process real-time signals that are critical to decisions related to electronic warfare and other tactical operations. Today, the need for intelligent systems with modern signal processing techniques is well felt. The main task of such systems is to identify the radars in the operating environment and classify them based on the previous learning of the system and perform the necessary operations at high speed and in real time, especially in cases where the received signal is related to an instantaneous threat such as missiles and electronic warfare systems. They may respond as a warning.The purpose of this study are to use the results of this research in classifying the information extracted by radar listening systems, which is achieved after the steps of selecting the input signal and selecting the correct classification algorithms, and another is to increase the speed using the vector vector digitization method. In this article, we present the data-driven methods of data collection using 4-digit vector learners and self-organizing methods.In this paper, we use learning vector quantization and self-organizing map methods to correlate the data. In this method, the neural network algorithm is first organized for the required coding positions, and in the next step, the quantization vector learning algorithm is created for data retrieval. In this article, we will also consider each database benchmark. The results obtained from the implementation of ordinary humanitarian command-and-control global standard deviation practices have been discussed in the light of the usual restraint methods, which demonstrate the great capability of these concepts.

Today, one of the most important challenges of information security and communication networks is the increasing number of malware and, consequently, finding suitable ways to protect systems against them. Knowing in time and finding ways to deal with the malicious effects of malware is one of the most important challenges for programmers and information security professionals. Is. Intelligent malware detection systems are able to model malicious behavior well. Extracting appropriate features and using efficient classifiers can improve the performance of such systems. In this paper, a new approach to malware detection is proposed using synergy of the features of the hardware counters and the optimization of the multilayer perceptron neural network classifier. The proposed system is able to identify healthy files from malware by extracting features with high discrimination and also using the neural network optimized by the dragonfly algorithm. In order to evaluate the proposed system, a data set including 168 healthy samples and 437 samples infected with malware is used. The results of the simulations show the higher performance of the proposed category compared to other categories, so that the proposed system has been able to detect the presence of malware-infected files with 86% accuracy.

Ensuring the security of the Internet of Things (IoT) services and the related applications is crucialin building users' trust in utilizing the Internet of Things platform. Data generated from varioussmart devices on the Internet of Things is one of the biggest concerns. Cloud computing hasemerged as a critical technology that can process such a large database repository of various devicesavailable on the Internet of Things. Authentication and privacy of IoT-enabled devices play acritical role in integrating the IoT and cloud computing technologies. The complexity and robustnessof authentication protocols are still the major challenges. This article provides a mutualauthentication protocol for IoT-enabled devices. AVISPA is used to evaluate the proposedprotocol's performance formally, and the MatLab tool is used to evaluate the time andcommunication costs. The results show the superiority of the proposed protocol compared to otherapproaches in terms of speed and robustness.

One of the dangers that is inflicted to the world by quantum algorithms is their usage to break the security of some systems based on classical cryptography. Nonetheless, the security of systems based on quantum cryptographic algorithms is safe from this risk. Another useful feature of quantum cryptography algorithms is the ability to figuring out any eavesdropping in the communication channel. Also, it is possible to safely distribute shares from the stock maker to the shareholders in the quantum secret sharing algorithm, while this is not possible in the classic secret sharing algorithm. One of the most important applications of the Internet of Things (IoT) network is the Military Internet of Things (MIoT). Due to the fact that MIoT has a direct impact on battlefields, military usage of IoT is more important than civilian applications. Therefore, security issues related to various parts of the MIoT are very critical. In this paper, we have presented a new design using a combination of classical cryptography (specifically the newly introduced digital signature scheme) and quantum cryptography (specifically quantum secret sharing) to improve the security of MIoT. Compared to similar research that have been conducted in this area, the introduced scheme provides more security requirements and is resistant to almost most of famous attacks on this network.

Compressed sensing (CS)-based inverse synthetic aperture radar (ISAR) imaging usually considers the uniform motion of targets. However, in practical scenarios, the targets usually have non-uniform motion, which creates the time-varying Doppler frequency shift and the ISAR image is blurred. Also, the basis matrix used in CS-based ISAR imaging is related to the rotational motion parameters which should be estimated too. However, the targets are assumed to have cooperative behavior with respect to radar; that is the target motion is known a priori and parameter estimation is not considered. In this paper, an improved version of CS-based imaging for non-uniform motion with constant acceleration and non-cooperative targets is proposed and best sparse representation is extracted. Simulation results show that the proposed algorithm is more efficient than other methods even without rotational motion compensation and provide higher image contrast.

Game theory uses mathematical models to analyze the methods of cooperation or competition of intelligent beings. Game theory attempts to model the mathematical behavior of strategic interaction among rational decision-makers. The ultimate goal of this knowledge is to find the optimal strategy for the players. One of the newest ideas in the application of Game theory in the field of artificial intelligence and machine learning is Generative Adversarial Networks. GANs consist of two parts, use Game theory and compete with each other, making it possible for unsupervised or semi-supervised learning. In addition to generating data, these networks are also used to identify malicious software and software security, machine translation, and natural language processing, and to build a three-dimensional model of an image. However, GANs have a very long training time due to the high number of epochs and input parameters. In this paper, in order to solve the problem of long training time of these networks in the classification of imbalanced high-dimensional datasets, a solution is presented that first, GAN-based oversampling on minority classes. Then in order to improve the efficiency of the designed GAN, the mentioned network is parallelized and ensemble classification is done. The different scenarios performed on the classification of diabetic retinopathy dataset by the proposed method.The results showed the classification accuracy of 87%, the training time is reduced by 74%, which shows higher accuracy than the latest scientific advances.

Inter-component communication and specifically pendingintents have expanded the development of android application. Although PendingIntent is used in many android applications, its improper use carries risks and can lead to various attacks such as denial of service, Privilege escalation and data leakage. Therefore, it is important to detect vulnerabilities associated with PendingIntent before android apps are published by Android app stores. One of the challenges of analyzing and detecting vulnerabilities for Android markets is the running duration of the vulnerability detection tools. In this paper, a new method has been proposed to detect vulnerabilities associated with PendingIntent. PIVATool is a tool based on static analysis for detecting PendingIntent-related vulnerabilities that takes less time to detect vulnerabilities without compromising precise. For evaluation, PIVATool is compared with PIAnalyzer tool. The results on 51 selected program benchmarks showed that PIVATool detects vulnerabilities on average 27% faster than PIAnalyzer with the same precise.

Security of electronic payment systems have become important, by increasing demand. Protecting against double spending, tracing malicious users, users' anonymity and privacy-preserving are important secure goals of any electronic payment systems. To achieve these goals, blockchain technology is useful, blockchain technology can solve many issues, such as bottlenecks, delays and operational risks that exist in the financial industry, but blockchain-based electronic payment systems cannot punish or trace malicious users without using trusted third party. In this article, we present a blockchain-based electronic payment scheme to protecting the anonymity and privacy of honest users, and also trace and punish malicious users with no need to TTP. We have used a fair blind digital signature scheme and a secret sharing scheme for this purpose. Users also use pseudonyms to maintain anonymity. users' pseudonyms are generated by pre-computations, therefore, the proposed scheme have a good performance.

Fuzzers can reveal vulnerabilities in the software by generating test input data and feeding inputs to software under test. The approach of grammar-based fuzzers is to search in the domain of test data which can be generated by grammar in order to find an attack vector with the ability to exploit the vulnerability. The challenge of fuzzers is a very large or infinite search domain and finding the answer in this domain is a hard problem. Grammatical Evolution(GE) is one of the evolutionary algorithms that can utilize grammar to solve the search problem. In this research, a new approach for generation of fuzz test input data by using grammatical evolution is introduced to exploit the cross-site scripting vulnerabilities. For this purpose, a grammar for generating of XSS attack vectors is presented and a fitness calculation function is proposed to guide the GE in search for exploitation. This method has realized the automatic exploitation of vulnerability with black-box approach. In the results of this research, 19% improvement achieved in the number of vulnerabilities discovered compared to the white-box method of NAVEX and black-box ZAP tool, and without any false positives.

Complex networks are currently being studied in many fields of science, and many natural systems can be described by them. The Internet and the brain, which are networks of routers and neurons, respectively, are examples of complex networks. There are also different types of complex networks, which can be referred to as scale free networks, small world networks and random networks. In this paper, an epidemic model of rumor spread in all three types of these networks is proposed. In this model, in addition to the existing cases (susceptible-infected-recovered), the rumor delay mechanism as well as the counter-attack mechanism have been added. The proposed model is presented as: Susceptible - Infected - Infected - Counterattack - Recovered - Susceptible (SECIRS). The methods of diffusion and decontamination for these three types of networks are compared. The simulation results are exactly in line with the theoretical analysis and show that in scale free networks, the spread of pollution is faster than the other two types. Pollution is lower and decontamination is faster.

The Non-orthogonal multiple access (NOMA) is known as a bandwidth efficient access method for fifth generation wireless systems. Unlike the conventional orthogonal multiple access method, which uses only one subcarrier to send information simultaneously, NOMA uses the power domain to serve multiple users from a number of subcarriers simultaneously. In previous attempts, the superiority of the non-orthogonal multiple access method has been proven to be the orthogonal access method of security-to-listening, which state by the Secrecy Sum Rate (SSR), but the superiority of the overall NOMA secrecy sum rate over the OMA, can’t be seen a lot of difference between them. The paper attempts to increase the secrecy sum rate by adding artificial noise along with the user's signal in the transmitter. In this way, legitimate users receive artificial noise signal and, as soon as they arrive, they remove this signal from the received signal, but the eavesdropper assume it is the signal of a legitimate user and they use a lot of energy to decode it. This will reduce the eavesdropper rate then reduce the listening probability. The results obtained in this paper show that, for example, in a specific SNR such as 10 dB, the total secrecy sum rate (SSR) for the orthogonal multiple access method is about 0.1, for the common non-orthogonal multiple access method is about 0.25 and for the non-orthodontic multiple access method with using the artificial noise is about 1 bit per second per hertz.

The importance of navigation precision in high dynamic environments has led to integrating the Inertial Navigation System (INS) with satellite navigation systems. In one of those integration methods that INS is integrated with GPS, GPS outage is an unavoidable challenge. Moreover, due to significant noisy signal existing in low-cost MEMS sensors, navigation precision severely decreases, and the INS error will diverge in the long term. This paper improves the INS/GPS navigation system using Artificial Intelligence (AI) during GPS outage. In this approach, the INS outputs at t and t-1 are injected to the AI module as the positioning and timing information. While GPS is available, the AI module is trained, and its output is compared with the GPS output. The AI module indeed intents to drive the INS output to the GPS output during GPS outage. To evaluate this approach and compare with some different intelligence systems, we have utilized Neural Networks (NNs) as an AI module in five different NNs: multilayer perceptron (MLP(, radial basis function )RBF(, support vector regression (SVR(, Wavelet, and adaptive neuro-fuzzy inference system (ANFIS). The required dataset to compare all five mentioned methods is gathered in a real environment by a mini-airplane. The results of all five methods represent that the proposed methods have superior performance compared to other traditional methods; so that the wavelet NN outperforms others by approximately 30%.

Today, many scientific problems require high computational power and storage space. Cloud computing is a model for easy access to different resources such as storage space with minimal need for service provider interaction. The cloud environment has been used for many benefits, but security and privacy issues are important challenges due to outsourcing. On the other hand, task scheduling is a fundamental issue in distributed systems such as cloud computing. Because there are several tasks to be performed that require different resources while resources are limited. Therefore, cloud tasks must be intelligently scheduled so that system performance and provider profitability are maximized. To solve this challenge, various techniques such as gradient-based algorithms for continuous and single-model problems are common. In cloud computing, due to the large search space and complex nature, these algorithms may not provide a suitable solution. Efficient meta-heuristic techniques can deal with these problems and find near-optimal solutions in a reasonable time. In this paper, a security-based scheduling algorithm using an improved Particle Swarm Optimization algorithm is presented. The improved algorithm uses multi adaptive learning to provide diversity in a population. Therefore, a good balance between exploration and exploitation. The proposed task scheduling algorithm simultaneously considers five parameters (i.e., round trip time, load, energy consumption, cost, and security) to provide load balancing and reduce energy consumption. The proposed algorithm is implemented using the CloudSim simulator and compared with the relevant strategies (i.e., CJS, OTSS, GTSA, and JSSS). The simulation results show that the proposed algorithm, considering the characteristics of tasks and resources, has significant efficiency and effectiveness in the cloud environment, especially at high workloads.

This paper presents a new approach to the detection of asymptomatic anomalies based on the signal processing related to local information of graph that simultaneously detects small compact anomalous subgraphs in the unknown graphs of large social networks. It also introduces a novell sampling algorithm based on compressive sensing to retrieve the sparse properties of static networks, which aims to improve the accuracy of anomaly detection while reducing the complexity of data sampling. The results of experimental experiments with artificial random and real datasets of social networks in comparison with the state-of-the-art methods showed that the proposed approach, in addition to having the accuracy of simultaneous detection of anomalous compact subgraphs, the computational complexity reduced from O(n^4 √(log⁡n )) to O(n^2) in the n node networks and is easily applicable in complex dynamic networks.

In wireless sensor networks (WSNs), the nodes around the sink deplete their energy considerably earlier than the other network nodes because they forward all the data packets toward the sink. Hence, the active nodes of the network and the sink are disconnected and also the network is divided to sub-networks. Utilizing mobile sink is a solution for this problem; using a mobile sink can balance the load of network but the manner of data transmission toward the mobile sink to have desired QoS and optimize delivery ratio, energy conservation, overhead and delay, is a challenge. In this paper, In this paper, a heuristic algorithm for data diffusion and gathering based on virtual line is proposed in WSNs with mobile sink. In the proposed algorithm, a virtual line is considered in the middle of the network and the rest of the network Area is zoned. The virtual line is segmented to facilitate the search phase. Also, the sensed data in different areas of the network are received only in the corresponding part of the virtual line. In this method, by sending sink requests in batches, reducing the data search area in the virtual line, as well as aggregating and reducing the final data volume, the data transfer operation to the mobile sink is performed. The evaluations show the proposed scheme overcomes the recent related works presenting more efficient energy conservation, delay, overhead and delivery ratio.