Research

Cybersecurity:

No matter how sophisticated and advanced an organization’s security system is, it remains vulnerable due to the human factor. In this paper, we conducted a survey to analyze the patterns used by the faculty, staff, and students when generating passwords at a small-sized university. We found that users are not as aware of security requirements and practices as they think. Moreover, the vast majority of users’ passwords are breakable within a few days or even less. Interestingly, we found that using numbers and uppercase letters is common among users. However, numbers are mostly used at the end of the passwords and uppercase letters are mostly used at the beginning of passwords. The existence of such trends makes it easier for attackers to generate more effective dictionaries. This project aims at analyzing password strength and hence making recommendations to IT personnel and the general public to strengthen the security of their passwords.

Image processing on-board spacecraft:

On-board autonomy is one of the key technologies for future spacecraft development. On-board image processing is an important component of future autonomous spacecraft. The main setback towards the on-board deployment of advanced image-processing algorithms is the high computational power requirements of such algorithms given the on-board computational constraints. The classification of satellite images provide increased on-board spacecraft autonomy which has become a strategic requirement for future space missions. Providing spacecraft with classification capabilities will enable advanced features such as the identification of regions of interest for priority transmission, autonomous mission rescheduling for priority image taking and generating summary products to conserve transmission bandwidth. Although there have been many literature sources on on-board image processing, there is no specific reference discussing the state-of-the-art and future trends of applying this important technology in small satellite missions.

 

Support Vector Machines on-board Small Satellites:

Support Vector Machines (SVM) are considered a relatively recent tool of computational intelligence. However, since their introduction in 1995, Support Vector Machines have found their way to various applications in engineering, economy and statistics. Comparative studies have shown that classification by a SVM can be more accurate than popular contemporary techniques such as neural networks and decision trees as well as conventional probabilistic classifiers such as the maximum likelihood classification. Despite their possible advantages, SVM use in space applications is still very limited, mainly because it is a relatively recent development when compared to its competitors.

The purpose of this project is to investigate the possible applications of SVM in space. This project proposes two specific space-based applications representing the two basic modes of SVM: classification and regression. The first proposed application utilizes SVM in classification mode in order to induce on-board intelligence and autonomy on-board spacecraft by providing spacecraft with the ability to identify high-priority image data such as natural disasters (floods, volcanoes, forest fires, etc.). Previously, this has been done through (pixel-by-pixel) change detection between two images, the new image with a previous image. In this research project we are proposing the use of SVM to classify images without the need of previous history. The project would investigate the feasibility of this approach by using a set of features as inputs to the SVM and analyzing the result of the output of the classification.

The second application proposed in this project is to make use of SVM in regression mode. The project would investigate using Support Vector Regression (SVR) in Fault Detection, Isolation and Recovery (FDIR) on-board. FDIR is critical for on-board autonomy. The SVR approach will be compared to the existing on-board FDIR approaches such as the Built-In Testing (BIT).

Generally speaking, machine learning is computationally demanding, which have always been a set-back for using them in space application. This project plans to overcome the computational demand hurdle through hardware implementations of SVM. In both of the proposed applications above, the computational demand in software will be analyzed. A hardware implementation will be implemented using FPGAs and its performance will be compared to the software implementation.

Publications:

Mohammed Awad, Zakaria Al-Qudah, Sahar Idwan, and Abdul Halim Jallad Password Security: Password Behavior Analysis at a Small University

Follow AURAK on:

Facebook   Tweeter   Youtube   Google Scholar   Instagram   Linked in   Location