An architecture framework for enhanced wireless sensor network security

Abstract

This thesis develops an architectural framework to enhance the security of Wireless Sensor Networks (WSNs) and provides the implementation proof through different security countermeasures, which can be used to establish secure WSNs, in a distributed and self-healing manner. Wireless Sensors are used to monitor and control environmental properties such as sound, acceleration, vibration, air pollutants, and temperature. Due to their limited resources in computation capability, memory and energy, their security schemes are susceptible to many kinds of security vulnerabilities. This thesis investigated all possible network attacks on WSNs and at the time of writing, 19 different types of attacks were identified, all of which are discussed including exposures to the attacks, and the impact of those attacks. The author then utilises this work to examine the ZigBee series, which are the new generation of wireless sensor network products with built-in layered security achieved by secure messaging using symmetric cryptography. However, the author was able to uniquely identify several security weaknesses in ZigBee by examining its protocol and launching the possible attacks. It was found that ZigBee is vulnerable to the following attacks, namely: eavesdropping, replay attack, physical tampering and Denial of Services (DoS). The author then provides solutions to improve the ZigBee security through its security schema, including an end-to-end WSN security framework, architecture design and sensor configuration, that can withstand all types of attacks on the WSN and mitigate ZigBee’s WSN security vulnerabilities.