<![CDATA[The development of IoT devices causes a change in many aspects of human life. Although this device has a limited resources, IoT devices can be used in every kind of environment. The use of IoT device in these environments make the security of IoT device important to study. One of biggest DoS attacks happen to IoT devices because there is no self-defense mechanism toward dangerous packets, so that IoT devices easily infected by Mirai botnet. A method choosen for this research to solve this problem is using Intrution Detection System(IDS). This IDS is expected to handle DoS attack in IoT devices with its limitation. Machine learning is chosen for detector in IDS because it's better for detecting anomalies, and also can run better in limited resources than other type of IDS. The Machine Learning algorithm is J48 because J48 has been prooven to detect anomaly better than other classification algorithms. There are few testing parameters used in this research; which are resource usage, detection engine accuracy, ability to give alert, logging ability, realibility in capturing packet in the network, and ability to handle the attack. Based on the evaluation results, this IDS can handle an attack, give alert, and do the logging process. This IDS is also able to classify the packet up to 100%, but this IDS has average 73.6% for capture packet from the network, so IDS can show alert in average of 17.42%. The resource usage in this IoT devices increases by average CPU usage 16% and memory usage 70MB. Based on these testing results, IDS can be used for solution to handle DOS attack in IoT devices.]]>
Copyrights © 2018
