Industrial Control Systems (ICSs), which work based on Wireless Sensor Networks (WSNs), are prone to hacking and attacks. In node simulation attacks against ICS networks, the enemy may capture a sensor node and then make multiple copies with the same identifier (ID), code, and encryption of the recorded node. Unfortunately, many Intrusion Detection Systems (IDSs) are not efficient to detect clone attacks in ICSs. An alternative solution to improve the performance of early detection is a honeypot. This paper proposes a centralized architecture for detecting copy or clone nodes using a local multicast intrusion detection system. We divide the WSN into sections and give each one an inspector node. Each inspector node monitors its region and uses the node ID to identify clone nodes. We offer solutions for situations where the cluster-head is endangered. We also provide solutions for other cases where the natural node is compromised. Our evaluations show that the proposed system maximizes the detection probability and, at the same time, has a low connection overhead. تفاصيل المقالة

Distributed Denial of Service (DDoS) attacks are a significant threat, especially for the Internet of Things (IoT). One approach that is practically used to protect the network against DDoS attacks is the honeypot. This study proposes a new adversarial Deep Reinforcement Learning (DRL) model that can deliver better performance using experiences gained from the environment. Further regulation of the agent's behavior is made with an adversarial goal. In such an environment, an attempt is made to increase the difficulty level of predictions deliberately. In this technique, the simulated environment acts as a second agent against the primary environment. To evaluate the performance of the proposed method, we compare it with two well-known types of DDoS attacks, including NetBIOS and LDAP. Our modeling overcomes the previous models in terms of weight accuracy criteria (> 0.98) and F-score (> 0.97). The proposed adversarial RL model can be especially suitable for highly unbalanced datasets. Another advantage of our modeling is that there is no need to segregate the reward function. تفاصيل المقالة