<![CDATA[Wireless Sensor Networks used to collect all of data and information in complex environments. WSN uses wireless communication channel technology to transmit data, without security protection, data is very vulnerable to internal and external attacks. Black hole is one of the attacks that occur on WSN. A black hole attack works by declaring itself to have the shortest route to the destination node. This is very dangerous if the package contains important information. From these problems, black hole attacks can be detected using the Support Vector Machine method on the Ad Hoc On-Demand Distance Vector reactive protocol. SVM is a powerful method for classification and regression. The purpose of this study is to determine the accuracy of SVM in detecting black hole attacks and how much data is discarded by these malicious nodes. In the test results, it can be seen when the simulation is under normal conditions, the packet loss is small (1), while in the simulation there are black holes (239) 100%. Simulation with normal conditions on the packet delivery ratio as much as 99,5816%, while the simulation has a black hole with a value of 0%. Simulations with normal conditions have an average end-to-end delay of 202,429 ms, while in the simulation there are black holes of about 172,524 ms to 214,979 ms. The results showed that the SVM algorithm in classifying black hole attacks with calculations produced accurate results.]]>
Copyrights © 2021
