الفهرس 
الغلافOnly 14 pages are availabe for public view
 |  | from | 110 |  |  |
| | | from | 110 | | |
Abstract
[1]. Our interest in this thesis is SQL injection. A SQL injection attack targets interactive web applications that employ database services. These applications accept user inputs and use them to form SQL statements at runtime [2].
A SQL injection attack consists of insertion or ”injection” of a SQL query via the input data from the client to the application. A successful SQL injection exploit can read sensitive data from the database, modify database data (In.sertlUpdatelDelete), execute administration operations on the database (such as shutdown the DBMS), recover the content of a given file existing on the DBMS file system and, in some cases, issue commands to the operating system. SQL injection attacks are a type of injection attack, in which SQL commands are injected into data-plane input in order to affect the execution ofpredefinedSQL commands [1].
This thesis presents a new model that addresses the problem of SQLiilj~ction of web applications. The proposed model called SIW (SQL injection Wall) which assists in detecting SQL injection attacks that may be happened in web applications or desktop applications. The proposed model combines static anlysis and runtime analysis of dynamically generated SQL queries to prevent SQL injection attack. SIW has the ability to easily expand without to cause any harming to any application that using it.
The developed model architecture consists of two main components SQL parser, Template Checker. The SQL parser component consists of two parts Syntax analyzer and Template Generator. Syntax analyzer parses two SQL statement strings; one SQL statement before user input and the other one after user input and passes them to Template Generator which generates two XML templates for the entered SQL statements. The other component
Template Checker compares two XML templates that were generated by Template Generator and decide whether SQL injection may occur or not.
The experiments results show that SIW can give best results in detecting SQL injection attacks. With 557 different user inputs; 537 vulnerable inputs that are used nowadays by SQL injection attackers which covers most types of SQL injection, and 20 legal inputs that contains special cases like special characters. SIW gives 96.59 % ratio of True behavior .. which means a successful classification of malicious inputs as attacks, and 3 Al % ratio of false behavior which means unsuccessful classification of legal inputs as legal or throws an exception. This result is efficient and is considered as good system behavior for different user inputs.