Desktop version

Home arrow Computer Science

  • Increase font
  • Decrease font

<<   CONTENTS   >>

An Effective Social Internet of Things (SIoT) Model for Malicious Node Detection in Wireless Sensor Networks


Internet of Things (IoT) has turned into a prevalent framework to help numerous advanced applications and services, for example, brilliant homes, smart healthcare, open security, modern observing, and condition assurance [1] to reinforce the unwavering quality and security in remote sensor systems. Hence, it is critical to structure a compelling security system for recognizing vindictive hubs in an IoT network [2]. IoT service enables certain capacities to be helped out through a predefined interface. A few researchers are especially keen on recognizing danger issues emerging during finding and coordinating information inside IoT [3]. The SIoT is a bigger social network, associating people and people and items, and articles and items. With numerous security issues in a SIoT model, it experiences a similar security problem as customary Internet-based as well as remote frameworks, including sticking, spoofing, and so forth [4]. The SIoT model is an interdisciplinary developing space that empowers self-ruling association among informal communication and the IoT. The nature of SIoTs presents different difficulties in its plan, design, execution, and activity of the executives [5]. SIoT has the ability to offer novel applications and networking services for IoT effectively [6].

Also, with the scope of social network service being extended from individual focused to a partnership focused, the association with IoT empowers a business practical joint effort [7]. The attackers change the conduct of the hubs in the system to fall and debase the usefulness of the wireless sensor networks. The malicious hubs in the wireless sensor systems can be identified utilizing cross-hybrid acknowledge scheme (HAS). In this strategy, the hubs in the wireless sensor system are gathered into a number of clusters. Each cluster ought to have just three hubs and an individual cluster key in all hubs in the cluster [8]. On the off chance that malicious hubs effectively alter the information, it can make an impact on the IoT work, that is, prompting an off-wrong decision [9].

A node is in the long run viewed as malicious when its total outcome progresses toward becoming lower than a specific limit. The disadvantage of this method is that it doesn’t identify replay assaults distinguished by optimization, machine learning, and deep learning procedures [10]. From these strategies, the researchers have derived system insights and malicious node practices and announced the effective discovery of particular sending, Hello flooding, and sticking assaults by their interruptions [11]. In light of these insights and practices, the malicious nodes are effectively and adequately confined to SIoT arrangement.


In 2019, Ande et al. [12] proposed the next-generation Internet, a web system that fuses human qualities. Security dangers inside all design layers and some mitigation methodologies are discussed with some future advancements. Given the potentially sensitive nature of IoT datasets, there is a need to build up a standard for the sharing of IoT datasets among the examination and professional networks and other important partners. The potential of blockchain technology is encouraging in secure sharing of IoT datasets Banerjee et al. [13]. A methodology, deep learning, to empower the discovery of assaults in social web of things has been presented by Diro et al. [14]. The deep learning model is thought about against the conventional machine learning approach, and the appropriated assault discovery is assessed against the brought-together detection system. The tests have demonstrated that our disseminated assault discovery system is better than the combined detection systems utilizing deep learning model.

The design is to arrange the assaults that don’t expressly harm the system; yet by contaminating the inside nodes, they are prepared to do the assaults on the system, which are named as inward assaults by Hajiheidari et al. [15]. At that point, categorizations of the IDSs in the IoT assault like denial of administration assault, Sybil assault, replay assault, particular sending assault, wormhole assault, black hole attack, sinkhole assault, sticking assault, false information assault have likewise been given utilizing regular highlights. The favorable circumstances and weaknesses of the chosen systems are also talked about. Parameter infusion, as a typical and incredible assault, is regularly abused by attackers to break into the servers of IoT by infusing malicious codes into the parameters of the solicitations, as explained by Yong et al. [16]. A hidden Markov model (HMM) based discovery framework is presented in this chapter, which is structured as a novel biregistry scoring design using both favorable and malevolent web traffic to protect against parameter infusion assaults in IoT systems.

<<   CONTENTS   >>

Related topics