There is a growing body of literature that recognizes the importance of data mining in educational systems. This recognition makes educational data mining a new growing research community. One way to achieve the highest level of quality in a higher education system is by discovering knowledge from educational data such as students' enrollment data. Many mining tools that aim to discover exciting correlations, frequent patterns, associations, or casual structures among sets of items in educational data sets have been proposed. One of the widely used tools is association rules. In this paper, the Apriori algorithm is used to generate association rules to discover the importance and correlation between factors that influence student's decision to enroll in higher education institutions in Sudan. The algorithm is applied using a student's enrollment data set that was created using a questionnaire and 800 students enrolled in governmental and private sector universities as a sample. This paper classifies factors that influence enrollment into: student's demographic factors and four categories of enrollment related factors (Student and Society, Educational Institution, Admission, and Employment related factors), and determines the most influential factors in determining student's decision to enroll in Sudanese universities. The analysis result shows that the Educational Institution related factors (50%) and Admission related factors (40%) are strongly influencing students' enrollment decision, while the Employment related factors (10%) and Student and Society related factors (0%) have weak influence. The factors out of the 14 Educational Institution related factors that have a high impact are: reputation, diversity of study, quality of education, education facilities, and feasibility.

The vulnerabilities of Mobile Ad hoc Networks (MANETs) make it subject to a large number of attacks. In order to understand the nature and behavior of such attacks, many classification schemes and taxonomies to MANET attacks have been proposed. This paper proposes a new taxonomy to MANETs attacks. The taxonomy is aimed to provide a consistent means of classifying attacks, as well as allowing previous knowledge to be applied to new attacks and providing a structured way to view such attacks. The taxonomy is based on attack attributes. Every attack is characterized by a unique vector of attribute values, where each attribute defines a specific attack property which may have different values. The taxonomy uses six attributes; the legitimacy of attacking node/s, the number of nodes participating in the attack, MANETs vulnerabilities utilized by the attack, the network resources exploited by the attacking node/s, the targeted victim and finally, the network security service compromised by the attack. The analysis of some well known attacks shows the capability of the proposed taxonomy in describing and categorizing these attacks as taxonomy vectors.

No abstract

No abstract

In this paper, we present a theoretical framework for a simple and efficient method that detects and blocks source IP spoofed packets and TCP/SYN flooding packets at source. The method is based on a network authentication server (AS), which performs an authentication process on SYN packets. The authentication process verifies the legitimacy of SYN packet's source IP address that initiate a connection request from a network subnet host to an external host. During the authentication process of SYN packets, AS identifies and blocks SYN packets with legal source IP address that chip in a TCP/SYN flooding attack. AS preserves network performance by exchanging authentication messages in plain text, and acts as a stateful inspection firewall and only SYN packets are subject for inspection. Our method which is capable to detect and prevent all types of spoofing packets including subnet spoofing contributes to standard ingress/egress methods in eliminating bogus traffic on the Internet.