No abstract

No abstract

No abstract

In recent years, due to the extensive use of the Internet, the number of networked computers has been increasing in our daily lives. Weaknesses of the servers enable hackers to intrude on computers by using not only known but also new attack-types, which are more sophisticated and harder to detect. To protect the computers from them, Intrusion Detection System (IDS), which is trained with some machine learning techniques by using a pre-collected dataset, is one of the most preferred protection mechanisms. The used datasets were collected during a limited period in some specific networks and generally don't contain up-to-date data. Additionally, they are imbalanced and cannot hold sufficient data for all types of attacks. These imbalanced and outdated datasets decrease the efficiency of current IDSs, especially for rarely encountered attack types. In this paper, we propose six machine-learning-based IDSs by using K Nearest Neighbor, Random Forest, Gradient Boosting, Adaboost, Decision Tree, and Linear Discriminant Analysis algorithms. To implement a more realistic IDS, an up-to-date security dataset, CSE-CIC-IDS2018, is used instead of older and mostly worked datasets. The selected dataset is also imbalanced. Therefore, to increase the efficiency of the system depending on attack types and to decrease missed intrusions and false alarms, the imbalance ratio is reduced by using a synthetic data generation model called Synthetic Minority Oversampling TEchnique (SMOTE). Data generation is performed for minor classes, and their numbers are increased to the average data size via this technique. Experimental results demonstrated that the proposed approach considerably increases the detection rate for rarely encountered intrusions.

Wireless Sensor Network (WSN) is a large scale network with from dozens to thousands tiny devices. Using fields of WSNs (military, health, smart home e.g.) has a large-scale and its usage areas increasing day by day. Secure issue of WSNs is an important research area and applications of WSN have some big security deficiencies. Intrusion Detection System is a second-line of the security mechanism for networks, and it is very important to integrity, confidentiality and availability. Intrusion Detection in WSNs is somewhat different from wired and non-energy constraint wireless network because WSN has some constraints influencing cyber security approaches and attack types. This paper is a survey describing attack types of WSNs intrusion detection approaches being against to this attack types.

No abstract

Solving the path planning problem of a UAV is a challenging issue especially if there are too many checkpoints to visit. Mainly, the brute force approach is used to find the shortest path in the mission area, which requires too many times to find a solution. Therefore, evolutionary algorithms and swarm intelligence techniques are used to find a feasible solution in an acceptable time. In this study, path planning problem of a UAV is solved by using a highly parallelized Ant Colony Optimization (ACO) algorithm on CUDA platform. The UAV path is constructed for disseminating keys and collecting data from a Wireless Sensor Network, which is previously defined.Due to its simplicity and effectiveness, ACO is selected as a path planning algorithm. However, ACO is not satisfactory if the mission area becomes large and there are an excessive number of checkpoints and/or additional constraints. In order to increase the performance, some parallelization techniques must be used in high performance computing platforms. GPU architecture has emerged as a powerful and low cost architecture for enabling impressive speedups for scientific calculations. Therefore, the parallel structure is constructed on CUDA architecture. The experimental results are compared with the CPU performance of the serial algorithm, and they clearly show that the proposed approach have a great potential for acceleration of ACO and allow to solve many complex tasks such as UAV path planning problem. We also present the execution results with different parameter values to expose the results for the researchers.

No abstract