Search

Routing Protocol for Low Power and Lossy Networks (RPL) is the de-facto routing standard in IoT networks. It enables nodes to collaborate and autonomously build ad-hoc networks modeled by tree-like destination-oriented direct acyclic graphs (DODAG). Despite its widespread usage in industry and healthcare domains, RPL is susceptible to insider attacks. Although the state-of-the-art RPL ensures that only authenticated nodes participate in DODAG, such hard security measures are still inadequate to prevent insider threats. This entails a need to integrate soft security mechanisms to support decision-making. This paper proposes iTRPL , an intelligent and behavior-based framework that incorporates trust to segregate honest and malicious nodes within a DODAG. It also leverages multi-agent reinforcement learning (MARL) to make autonomous decisions concerning the DODAG. The framework enables a parent node to compute the trust for its child and decide if the latter can join the DODAG. It tracks the behavior of the child node, updates the trust, computes the rewards (or penalties), and shares them with the root. The root aggregates the rewards/penalties of all nodes, computes the overall return, and decides via its ϵ -Greedy MARL module if the DODAG will be retained or modified for the future. A simulation-based performance evaluation demonstrates that iTRPL learns to make optimal decisions with time. [ABSTRACT FROM AUTHOR]

Due to the properties of ad-hoc networks, it appears that designing sophisticated defence schemes with more computing capital is impossible in most situations. Recently, an inconsistency in the ad-hoc design of intrusion detection in the network has gotten a lot of coverage, with these intrusion detection techniques operating in either cluster-based or host-based configurations. The host and cluster-based systems have advantages and disadvantages, such as the network preserve security in case of delay in replacing a cluster head. Many detection systems in these networks use a supervised learning method to learn from shared routing knowledge. Deep learning is the trending supervised learning method which is been suggested for many applications, due to its deep feature extraction and classification capability. The deep learning method is best suitable to resolve the problems of the ad hoc network. But due to its limitation of supervised learning nature, more research finds are needed before implementation. These intelligence methods need a massive labeled dataset to self-train and take a decision in real-time. Also, these methods will be vulnerable to new attacks. To address the issues posed, the deep learning approach requires a technique of incorporating unsupervised learning behaviours. This paper proposes and highlights the methodology - Deep Embedded Median Clustering (DEMC), which performs two-phase operations (1) Organization of latent feature space (2) K-median clustering to cluster the Z with Kullback–Leibler divergence as the objective function. Many researchers suggested various methodologies for better anomaly detection in the network, but the knowledge gap and the possibilities for a better solution still exist. This study explores the new possibility and potential of an unsupervised learning technique that works with the nature of deep learning for analyzing and detecting anomalies and intrusion in ad hoc networks. The test to check the DEMC ability has been organized, and the findings are tabulated for analysis. [ABSTRACT FROM AUTHOR]