Key management for wireless sensor network security

Abstract

Wireless Sensor Networks (WSNs) have attracted great attention not only in industry but also in academia due to their enormous application potential and unique security challenges. A typical sensor network can be seen as a combination of a number of low-cost sensor nodes which have very limited computation and communication capability, memory space, and energy supply. The nodes are self-organized into a network to sense or monitor surrounding information in an unattended environment, while the self-organization property makes the networks vulnerable to various attacks.Many cryptographic mechanisms that solve network security problems rely directly on secure and efficient key management making key management a fundamental research topic in the field of WSNs security. Although key management for WSNs has been studied over the last years, the majority of the literature has focused on some assumed vulnerabilities along with corresponding countermeasures. Specific application, which is an important factor in determining the feasibility of the scheme, has been overlooked to a large extent in the existing literature.This thesis is an effort to develop a key management framework and specific schemes for WSNs by which different types of keys can be established and also can be distributed in a self-healing manner; explicit/ implicit authentication can be integrated according to the security requirements of expected applications. The proposed solutions would provide reliable and robust security infrastructure for facilitating secure communications in WSNs.There are five main parts in the thesis. In Part I, we begin with an introduction to the research background, problems definition and overview of existing solutions. From Part II to Part IV, we propose specific solutions, including purely Symmetric Key Cryptography based solutions, purely Public Key Cryptography based solutions, and a hybrid solution. While there is always a trade-off between security and performance, analysis and experimental results prove that each proposed solution can achieve the expected security aims with acceptable overheads for some specific applications. Finally, we recapitulate the main contribution of our work and identify future research directions in Part V.