Trust-Based Analytical Models for Secure Wireless Sensor Networks

Trust-Based Analytical Models for Secure Wireless Sensor Networks

Aminu Bello Usman (Auckland University of Technology, New Zealand) and Jairo Gutierrez (Auckland University of Technology, New Zealand)
Copyright: © 2018 |Pages: 19
DOI: 10.4018/978-1-5225-5583-4.ch002

Abstract

In this chapter, the authors hypothesize that in the design of a trust-based routing protocol, the exploration of the peers' routing attributes could significantly improve trust evaluation accuracy. In this regard, they study the properties of complex networks and their impact on trust and reputation propagation and evaluation. They start by illustrating the structural transitivity in the network and its approximation. They then proceed to present the theoretical and analytical relationship between trust and reputation model accuracy, average structural transitivity between peers, average shortest path between peers, and energy consumed by peers for trust and reputation propagation and evaluations. The experimental studies using simulation have further supported the results of the analytical study. In this chapter, the authors are paving a new angle of research on exploring the complex network properties impact on trust and reputation evaluation between wireless peers.
Chapter Preview
Top

Introduction

The security in computing domain aim at guarding the information or protecting an information system through a range of security policies, strategies, security products, and cryptographic techniques (Pathan, 2016), and the goal of security in the network is to protect the information and systems’ integrity, confidentiality and availability of data. However, in wireless Peer-to-Peer (P2P) networks this is a challenging goal. These networks have no central authority to perform the processes of authentication and verification of the registered peers in the network, and there are no central and high processing control points to perform the cryptographic computation and cryptographic key management that are required. Under those circumstances, the decision-making ability of the peers can result on the adoption of a selfish behavior. For example, how can a peer know whether its communication partner is genuine or malicious? Can the request, content or message from any neighboring device be trusted? How can the peers perform self-organization processes to provide a secure network paradigm with a high quality of service? Obviously, to answer these questions, a distributed mechanism is needed to ensure a secure decentralized P2P network system.

Over the years, researchers have resorted to going back to the drawing board to borrow the concept of trust as an alternative strategy for addressing security problems in P2P networks (Boukerch, Xu, & El-Khatib, 2007). The idea behind trust and reputation models is for the peers to rate each other and then use the aggregated ratings to derive the trust scores, which can assist the peers in deciding whether to collaborate or not to collaborate in the future tasks (Bello, Liu, Bai, & Narayanan, 2015a). Many authors in the literature on trust and reputation in various fields show that the formation of effective cooperation, norms, and trust in the network largely depend on the type of actors, and relationships between the actors involved (Cho, Swami, & Chen, 2011). This is consistent with the suggestions proposed in (Chen & Cai, 2005) and indeed other relevant work, on the factors that influence efficient communication between the peers in P2P networks: 1) Consideration of the type of peers interaction such as: Human-to-human, Machine-to-machine, and Human-to-machine and 2) the behaviour of each peer in the network.

The study of trust and reputation was originally explored in sociology and other social sciences fields of study. Most the literature and the studies of trust in sociology pay more attention to exploring and debating on the natural factors that determine and influence trust including: relationship between actors, strong and weak ties between peers, structural properties which are fundamentals and natural attributes that can be used in predicting trust and reputation behaviour of the actors in the community. For a decade, the concept of trust and reputation became one of the dominant research topics in computer science, and information technology due to their significance to many related disciplines. Recently, due to the rapid development of complex and autonomous P2P networks, the concepts of trust, reputation and cooperation received enormous attention in the field of peer-to-peer (P2P) security. In addition, the concept of trust and reputation schemes was explored by different researchers due to the ability of trust and reputation to enable peers to cope with the uncertainty and uncontrollability caused by the free will of others’ peers in the network.

Most of the literature and the studies on trust usually define the term based on their disciplines’ discretion and views. For example, from the social psychological and sociological views, the work of (Lusher, Robins, Pattison, & Lomi, 2012) define trust as a unidirectional relation between trustor and a trustee expressing the firm belief of the trustor that the trustee will behave as expected with respect to a particular level of reliability within a specific context and at a particular time; further, the work defines trust as a social construct and it discusses the natural attributes of the relationship among social actors (group or individual), furthermore, (Grandison & Sloman, 2000) define trust as the subjective belief of someone in the character, ability, strength, reliability, honesty or truth of someone or something.

Complete Chapter List

Search this Book:
Reset