Journal of Cyber Security and Mobility
ISSN / EISSN : 2245-1439 / 2245-4578
Published by: River Publishers (10.13052)
Total articles ≅ 261
Latest articles in this journal
Journal of Cyber Security and Mobility pp 1–28-1–28; https://doi.org/10.13052/jcsm2245-1439.1111
Almost all spatial domain image steganography methods rely on modifying the Least Significant Bits (LSB) of each pixel to minimize the visual distortions. However, these methods are susceptible to LSB blind attacks and quantitative steganalyses. This paper presents an adaptive spatial domain image steganography algorithm for hiding digital media based on matrix patterns, named “Adaptive Matrix Pattern” (AMP). The AMP method increases the security of the steganography scheme of largely hidden messages since it adaptively generates a unique codebook matrix pattern for each ASCII character in each image block. Therefore, each ASCII character gets a different codebook matrix pattern even in different regions of the same image. Moreover, it uses a preprocessing algorithm to identify the most suitable image blocks for hiding purposes. The resulting stego-images are robust against LSB blind attacks since the middle bits of green and blue channels generate matrix patterns and hiding secrets, respectively. Experimental results show that AMP is robust against quantitative steganalyses. Additionally, the quality of stego-images, based on the peak signal-to-noise ratio metric, remains high in both stego-RGB-image and in the stego-blue-channel. Finally, the AMP method provides a high hiding capacity, up to 1.33 bits per pixel.
Journal of Cyber Security and Mobility pp 745–774-745–774; https://doi.org/10.13052/jcsm2245-1439.1046
Over the past decade, digital communication has reached a massive scale globally. Unfortunately, cyberbullying has become prevalent, with perpetrators hiding behind the mask of relative internet anonymity. In this work, efforts were made to review prominent classification algorithms and also to propose an ensemble model for identifying cases of cyberbullying, using Twitter datasets. The algorithms used for evaluation are Naive Bayes, K-Nearest Neighbors, Logistic Regression, Decision Tree, Random Forest, Linear Support Vector Classifier, Adaptive Boosting, Stochastic Gradient Descent and Bagging classifiers. Through experimentations, comparisons were made with the classifiers against four metrics: accuracy, precision, recall and F1 score. The results reveal the performances of all the algorithms used with their corresponding metrics. The ensemble model generated better results while Linear Support Vector Classifier (SVC) was the least effective of all. Random Forest classifier has shown to be the best performing classifier with medians of 0.77, 0.73 and 0.94 across the datasets. The ensemble model has shown to improve the results of its constituent classifiers with medians of 0.77, 0.66 and 0.94, as against the 0.59, 0.42 and 0.86 of Linear Support Vector Classifier.
Journal of Cyber Security and Mobility pp 725–744-725–744; https://doi.org/10.13052/jcsm2245-1439.1045
The bottleneck of all cryptosystems is the difficulty of the computational complexity of the polynomials multiplication, vectors multiplication, etc. Thus most of them use some algorithms to reduce the complexity of the multiplication like NTT, Montgomery, CRT, and Karatsuba algorithms, etc. We contribute by creating a new release of NTRUencrypt1024 with great improvement, by using our own polynomials multiplication algorithm operate in the ring of the form Rq=Zq[X]/(XN+1), combined to Montgomery algorithm rather than using the NTT algorithm as used by the original version. We obtained a good result, our implementation outperforms the original one by speed-up of a factor up to (X10) for encryption and a factor up to (X11) for decryption functions. We note that our improved implementation used the latest hash function standard SHA-3, and reduce the size of the public key, private key, and cipher-text from 4097 bytes to 2049 bytes with the same security level.
Journal of Cyber Security and Mobility pp 641–662-641–662; https://doi.org/10.13052/jcsm2245-1439.1041
The physical breach across the borders is a very common issue these days among nations sharing boundaries. It is controlled via proper border surveillance system. The border surveillance system is trivially a physical border intrusion detection system in which CCTV cameras are used traditionally to observe manually the presence of some intruder. Instead, we utilize the raspberry PI controller board based wireless sensor nodes fitted with raspberry PI camera for identifying the intruder. Once the intruder is identified, the wireless sensor nodes communicate the messages with the next hop sensor nodes and the message ultimately reaches the control room from where army action may be initiated. In this work, we propose a novel lightweight security scheme (LSS) for raspberry PI based wireless node communication for the Border Surveillance System. We have utilized the XBee (Zigbee) serial communication between raspberry PI based wireless sensor nodes. The proposed scheme is based upon the notion of confusion and correct identification of pattern (byte) in the transmitted messages. The entire communication scheme is lightweight and secure.
Journal of Cyber Security and Mobility pp 699–724-699–724; https://doi.org/10.13052/jcsm2245-1439.1044
The advancement of information communication technology has triggered a revolution in using the Internet for legitimate educational purposes on university campuses. Therefore, the Internet has changed the way of human communication and contributed to the development of mankind. On the other hand it is regrettable that its revolution has helped malicious users to exploit it for the malign purpose to commit a cyberspace crime that has in turn negatively affected fellow users who were preyed on by cyber predators. This work aimed to examine the awareness of cybersecurity, the measures taken to protect against cyberattacks and the state of victimization among professors at Ambo University. Thus, the present study comes up with the following findings. First, the result shows that the respondents’ cybersecurity awareness was significantly influenced by cyber-crime victimization, fields of study, and protection measures. Second, the current study also depicts that the respondents’ protection measures were connected to and influenced by cyber-crime victimization, education level, and cyber-security awareness. Finally, the study’s findings show that being a cyber-crime victim has been linked to predictors’ variables: protection measures and the level of cybersecurity awareness.
Journal of Cyber Security and Mobility pp 679–698-679–698; https://doi.org/10.13052/jcsm2245-1439.1043
Content Delivery Networks (CDN) are the backbone of Internet. A lot of research has been done to make CDNs more reliable. Despite that, the world has suffered from CDN inefficiencies quite a few times, not just due to external hacking attempts but due to internal failures as well. In this research work the authors have analyzed the performance of a content delivery network through various reliability measures. Considering a basic CDN workflow they have calculated the reliability and availability of the proposed multi-state system using Markov process and Laplace transformation. Software/Hardware failures in any network component can affect the reliability of the whole system. Therefore, the authors have analyzed the obtained results to find major causes of failures in the system, which when avoided, can lead to a faster and more efficient distribution network.
Journal of Cyber Security and Mobility pp 663–678-663–678; https://doi.org/10.13052/jcsm2245-1439.1042
With the increase in the discovery of vulnerabilities, the expected exploits occurred in various software platform has shown an increased growth with respect to time. Only after being discovered, the potential vulnerabilities might be exploited. There exists a finite time lag in the exploitation process; from the moment the hackers get information about the discovery of a vulnerability and the time required in the final exploitation. By making use of the time lag approach, we have developed a framework for the vulnerability exploitation process that occurred in multiple stages. The time lag between the discovery and exploitation of a vulnerability has been bridged via the memory kernel function over a finite time interval. The applicability of the proposed model has been validated using various software exploit datasets.
Journal of Cyber Security and Mobility pp 569–592-569–592; https://doi.org/10.13052/jcsm2245-1439.1034
The increase in the deployment of IOT networks has improved productivity of humans and organisations. However, IOT networks are increasingly becoming platforms for launching DDOS attacks due to inherent weaker security and resource-constrained nature of IOT devices. This paper focusses on detecting DDOS attack in IOT networks by classifying incoming network packets on the transport layer as either “Suspicious” or “Benign” using unsupervised machine learning algorithms. In this work, two deep learning algorithms and two clustering algorithms were independently trained for mitigating DDOS attacks. Emphasis was laid on exploitation based DDOS attacks which include Transmission Control Protocol SYN-Flood attacks and UDP-Lag attacks. Mirai, BASHLITE and CICDDOS2019 datasets were used in training the algorithms during the experimentation phase. The accuracy score and normalized-mutual-information score are used to quantify the classification performance of the four algorithms. Our results show that the autoencoder performed overall best with the highest accuracy across all the datasets.
Journal of Cyber Security and Mobility pp 593–616-593–616; https://doi.org/10.13052/jcsm2245-1439.1035
Empirical channel models were always an important tool for proper wireless network planning. These models consider the properties of electromagnetic waves and terrain conditions. The efficiency and accuracy of these Empirical models suffer when they are used for an area other than where they have been designed. So tuning of these models is required for proper and accurate prediction of coverage and it is done by taking the correction factor into account. Comparison of four Empirical models i.e, the Lee, the ECC-33 model, the WI model, the Ericsson model, the COST 231, and SUI is done with Measured path loss, and the best model with minimum error is then selected for tuning. Field data of LTE network at 2300 MHz is collected at two sites of Uttarakhand-India. It is analysed that the Ericsson model shows minimum RMSE, Standard Deviation, and Mean error as compared to measured path loss, followed by the Okumura model. The Ericsson model is then tuned to further reduce the error. Validation of the tuned model is done at Haridwar.
Journal of Cyber Security and Mobility pp 617–640-617–640; https://doi.org/10.13052/jcsm2245-1439.1036
Malicious software, called malware, can perform harmful actions on computer systems, which may cause economic damage and information leakage. Therefore, malware classification is meaningful and required to prevent malware attacks. Application programming interface (API) call sequences are easily observed and are good choices as features for malware classification. However, one of the main issues is how to generate a suitable feature for the algorithms of classification to achieve a high classification accuracy. Different malware sample brings API call sequence with different lengths, and these lengths may reach millions, which may cause computation cost and time complexities. Recurrent neural networks (RNNs) is one of the most versatile approaches to process time series data, which can be used to API call-based Malware calssification. In this paper, we propose a malware classification model with RNN, especially the long short-term memory (LSTM) and the gated recurrent unit (GRU), to classify variants of malware by using long-sequences of API calls. In numerical experiments, a benchmark dataset is used to illustrate the proposed approach and validate its accuracy. The numerical results show that the proposed RNN model works well on the malware classification.