| Title | Predicting Cyber Attacks in a Proxy Server using Support Vector Machine (SVM) Learning Algorithm |
| Publication Type | Conference Paper |
| Year of Publication | 2021 |
| Authors | Agana, Moses Adah, Edu, Joseph Ikpabi |
| Conference Name | 2021 IST-Africa Conference (IST-Africa) |
| Keywords | composability, Computers, DoS, learning, machine, Prediction algorithms, Predictive Metrics, Programming, proxy, pubcrawl, Resiliency, security, server, Servers, Support, Support vector machines, Task Analysis, vector |
| Abstract | This study used the support vector machine (SVM) algorithm to predict Denial of Service (DoS) and Distributed Denial of Service (DDoS) attacks on a proxy server. Proxy-servers are prone to attacks such as DoS and DDoS and existing detection and prediction systems are inefficient. Three convex optimization problems using the Gaussian, linear and non-linear kernel methods were solved using the SVM module to detect the attacks. The SVM module and proxy server were implemented in Python and javascript respectively and made to run on a local network. Four other computers running on the same network where made to each communicate with the proxy server (two dedicated to attack the server). The server was able to detect and filter out the malicious requests from the attacking clients. Hence, the SVM module can effectively predict cyber attacks and can be integrated into any server to detect such attacks for improved security. |
| Citation Key | agana_predicting_2021 |
Predicting Cyber Attacks in a Proxy Server using Support Vector Machine (SVM) Learning Algorithm