International Journal of Computer Networks and Communications Security

Volume 4, Issue 1, January 2016

 

 

 

AdaBoost Ensemble Learning Technique for Optimal Feature Subset Selection
Pages: 1-11 (11) | [Full Text] PDF (476 KB)
SOM KAMEL, NH HEGAZI, HM HARB, ASTE DEIN, HHME KADER
Electronic Research Institute, Department of Informatics
Faculty of Engineering in Azahr University, Department of Computer Science
Faculty of Engineering in Benha, Department of Communication

Abstract - As a result of the spread of technology in the world, it must be secured to the communication data. In the recent year, the trend of various network organizations is the maintaining a high level of security to get the secure data communication. Data communication over the internet is exposed to a huge number of threats where the intrusion prevention system (IPS) can be used to detect many threats in real time. The classification of network events is an important part in IPS to detect threats. IPS performance is based on the dimensionality reduction of features. The feature selection is widely used in data mining approach. The proposed experiment identifies the best selected features by using Best First search technique in the wrapper model to obtain better performance of IPS. The proposed system shows the novel AdaBoost ensemble learner algorithm which consists of the number of the base learner algorithm. The base learner algorithm is learning Bayesian Network by using Genetic Algorithm global search approach to reduce the classification time of AdaBoost ensemble learner and improve the performance of BN. The proposed system consists of five stages; training dataset pre-processing, subset generation, model validation, model evaluation, and model comparison which will apply seven classification algorithms to evaluate on the training dataset to detect threats.
 
Index Terms - Intrusion Prevention System (IPS), Feature Selection, Wrapper Model, AdaBoost, Genetic Algorithm

Citation - SOM KAMEL, NH HEGAZI, HM HARB, ASTE DEIN, HHME KADER. "AdaBoost Ensemble Learning Technique for Optimal Feature Subset Selection." International Journal of Computer Networks and Communications Security 4, no. 1 (2016): 1-11.
 

ASQL: A New Approach for Resource Auto-Scaling Using Q-Learning in Cloud Computing Environment
Pages: 12-20 (9) | [Full Text] PDF (550 KB)
B Asgari, MG Arani
Department of Computer Engineering, Mahallat Branch, Islamic Azad University, Mahallat, Iran
Department of Computer Engineering, Parand Branch, Islamic Azad University, Tehran, Iran

Abstract - Cloud services have become more popular among users these days. Providing automatic resource for cloud services is one of the important challenges. In cloud computing environment, resource providers shall offer required resources to users automatically without any limitations. It means whenever a user needs more resources, the required resources should be dedicated to the users without any problems. On the other hand if resources are more than users needs extra resources should be turn off temporarily and turn back on whenever they needed. In this article the approach is to represent enforcement learning-aware for auto-scaling resources according to Markov decision process (MDP). Results would show the rate of SLA (service level agreement) violation and stability that proposed better functions compared to the similar approaches.
 
Index Terms - Cloud Computing, Scalability, Auto-Scaling, Reinforcement Learning

Citation - B Asgari, MG Arani. "ASQL: A New Approach for Resource Auto-Scaling Using Q-Learning in Cloud Computing Environment." International Journal of Computer Networks and Communications Security 4, no. 1 (2016): 12-20.
 

Bring Your Own Device: Security Challenges and A theoretical Framework for Two-Factor Authentication

Pages: 21-32 (12) | [Full Text] PDF (435 KB)
M OLALERE, MT ABDULLAH, R MAHMOD, A ABDULLAH
Department of Cyber Security Science, Federal University of Technology Minna, Nigeria
Information Security Research Group, Faculty of Computer Science and Information Technology, Universiti Putra Malaysia, Malaysia

Abstract - In this paper, the security challenges of BYOD are discussed, including existing security solutions which often are too restrictive. Data leakage is one of the security challenges confronting BYOD. Data leakage can occur as a result of stolen, lost or compromised employee devices. When an employee device is stolen, lost or comprised, an attacker can obtain access directly to the enterprise data on the employee device if a strong authentication technique is not in place. The traditional means of authenticating employees when connecting to an enterprise server in a traditional network environment which relies on either knowledge or ownership is too weak for the BYOD environment. In such a traditional enterprise network, employees obtain access to an enterprise server using their respective stationary desktop, while in a BYOD environment access to an enterprise server is from anywhere, making it easy for an attacker in possession of an employee device and password to gain unauthorised access. To address this problem, there is need for a strong authentication technique. This study proposes a theoretical framework for a two-factor authentication method that combines knowledge-based (Password) and biometric-based (Keystroke dynamic) features for authentication of mobile devices in a BYOD environment. Technical details on how the framework can be implemented are presented. It is the belief of the authors that proper implementation of the proposed potential future application framework will go a long way in addressing the problem of data leakage in a BYOD environment.
 
Index Terms - BYOD, Mobile Device, Authentication, Biometric, Keystroke Dynamic

Citation - M OLALERE, MT ABDULLAH, R MAHMOD, A ABDULLAH. "Bring Your Own Device: Security Challenges and A theoretical Framework for Two-Factor Authentication ." International Journal of Computer Networks and Communications Security 4, no. 1 (2016): 21-32.