Scalable, adaptive and survivable trust management for community of interest based Internet of Things systems

Abstract

Abstract-An Internet of Things (IoT) system connects a large amount of tags, sensors, and mobile devices to facilitate information sharing, enabling a variety of attractive applications. It challenges the design and evaluation of IoT systems to meet the scalability, compatibility, extendibility, dynamic adaptability and resiliency requirements. In this paper, we design and evaluate a scalable, adaptive and survivable trust management protocol in dynamic IoT environments. Recognizing that entities in an IoT system are connected through social networks of entity owners, we consider a community of interest (CoI) based social IoT where nodes form into communities of interest. Given inter-CoI vs. intra-CoI social connections among entity owners as input, we identify best trust protocol settings for achieving convergence, accuracy, dynamic adaptability and resiliency properties in the presence of dynamically changing conditions and malicious nodes performing trust-related attacks. For scalability, we consider a design by which a node only keeps trust information of a subset of nodes meeting its interest and performs minimum computation to update trust. We validate our design by extensive simulation considering both limited and ideal (unlimited) storage space. The results demonstrate that our trust management protocol using limited storage space achieves a similar performance level compared with the one under ideal storage space, and a newly joining node can quickly build up trust towards other nodes with desirable accuracy and convergence behavior.

Trust Management in Social Internet of Things: A Survey

“…In [5], the authors aim at improving the trust management protocol proposed in [4]. They reuse the same trust measure, which is an aggregation of honesty, cooperativeness and community-interest.…”

A Comparative Analysis of Scalable and Context-Aware Trust Management Approaches for Internet of Things

“…[1] notes that the IoT assists applications such as continuous e-health and smart product management by capturing, processing and sharing data, which necessitates the use of effective trust management protocols to manage trust between different IoT entities. However, [11] argues that trust management is constrained by the vast quantity of IoT entities, which challenge scalability with respect to accommodating the growing number of computational and storage entities.…”

“…This implies that trust management protocols for IoT networks should be highly resilient to trust-based attacks to endure security issues in hostile environments. According to [1], scalability should be a key consideration in the design of trust/reliable management protocols for IoT. In other works, the trust management protocols proposed by [2], [5], [8], [18], and [12] did not address scalability, undermining their applicability in large-scale IoT networks.…”

“…We now outline and discuss the working of each of these methods. Firstly, [1] proposed the Scalable, Adaptive and Survivable Trust Management for Community of Interest (CoI) based IoT, recognizing that nodes in IoT networks are owned by individuals and interconnected by social networks. To achieve scalability, they designed a protocol whereby each node can store the trust relationship data of a set of nodes within its CoI, thus enhancing convergence.…”

“…In addition, it would be possible to identify things that require replacement, repair, review, or that are obsolete [12]. Trust management protocols were proposed by [17], [2], [1], and [5] that did not address the context awareness issue. [9] argues that stakeholders in the IoT area of mobile, wearable and ubiquitous computing have recognized the need to secede from the conventional desktop model as more and more devices become mobile.…”