Detections of IoT Attacks via Machine Learning-Based Approaches with Cooja

Once hardware becomes "intelligent", it is vulnerable to threats. Therefore, IoT ecosystems are susceptible a variety of attacks and considered challenging due heterogeneity dynamic ecosystem. In this study, we proposed method for detecting that based on ML-based approaches release the final decision detect attacks. However, have implemented three as sample in via Contiki OS generate real dataset IoT-based features containing mix data from malicious nodes normal network be utilized models. As result, multiclass random forest model achieved 98.9% overall accuracy novel compared tree jungle, regression, boosted which 87.7%, 93.2%, 87.1%, respectively. Thus, tree-based approach efficiently manipulates analyzes KoÜ-6LoWPAN-IoT dataset, generated Cooja simulator, inconsistent behavior classify activities.