Thesis
Prevention and detection of distributed denial of service (DDoS) attacks using estimation and machine learning techniques
Washington State University
Master of Science (MS), Washington State University
2011
Handle:
https://hdl.handle.net/2376/103993
Abstract
A distributed denial of service (DDoS) attack is a threat to the availability of a computer resource and prevents its normal functioning by overwhelming the system with huge amount of requests. The target of such an attack is normally a web server belonging to vendors such as banks, online retailers etc. The perpetrators launch the attack to make the system unavailable to the legitimate users causing loss of money and goodwill. This attack has been under observation for quite some time, however the solutions provided so far focuses on a specific type of DDoS attack. Owing to the polymorphic nature of the attack, this proposed solution provides a generic method of detection that involves intelligence in the form of machine learning algorithms to detect all types of DDoS attacks. The main idea is to prevent the system from being flooded with a large number of requests than it can handle. The proposed solution has been categorized into estimation, feature calculation and traffic analysis modules. The estimation module calculates and predicts the number of requests arriving at the web server. If the number of requests is close to the pre-calculated threshold of the system the detection module containing machine learning algorithms is alerted. The learning algorithm analyses the traffic and checks for presence of rogue requests by observing characteristics and calculating features of the incoming traffic. Therefore, the system under consideration is proiv tected from the large number of requests well before the resources are exhausted, thus preventing the distributed denial of service attack.
Metrics
30 File views/ downloads
26 Record Views
Details
- Title
- Prevention and detection of distributed denial of service (DDoS) attacks using estimation and machine learning techniques
- Creators
- Smita Kamath
- Contributors
- Min Sik Kim (Degree Supervisor)
- Awarding Institution
- Washington State University
- Academic Unit
- Electrical Engineering and Computer Science, School of
- Theses and Dissertations
- Master of Science (MS), Washington State University
- Publisher
- Washington State University; Pullman, Wash. :
- Identifiers
- 99900525399001842
- Language
- English
- Resource Type
- Thesis