Robust data protection and high efficiency for IoTs streams in the cloud

Abuadbba, A 2017, Robust data protection and high efficiency for IoTs streams in the cloud, Doctor of Philosophy (PhD), Science, RMIT University.


Document type: Thesis
Collection: Theses

Attached Files
Name Description MIMEType Size
Abuadbba.pdf Thesis application/pdf 2.09MB
Title Robust data protection and high efficiency for IoTs streams in the cloud
Author(s) Abuadbba, A
Year 2017
Abstract Remotely generated streaming of the Internet of Things (IoTs) data has become a vital category upon which many applications rely. Smart meters collect readings for household activities such as power and gas consumption every second - the readings are transmitted wirelessly through various channels and public hops to the operation centres. Due to the unusually large streams sizes, the operation centres are using cloud servers where various entities process the data on a real-time basis for billing and power management. It is possible that smart pipe projects (where oil pipes are continuously monitored using sensors) and collected streams are sent to the public cloud for real-time flawed detection. There are many other similar applications that can render the world a convenient place which result in climate change mitigation and transportation improvement to name a few. Despite the obvious advantages of these applications, some unique challenges arise posing some questions regarding a suitable balance between guaranteeing the streams security, such as privacy, authenticity and integrity, while not hindering the direct operations on those streams, while also handling data management issues, such as the volume of protected streams during transmission and storage. These challenges become more complicated when the streams reside on third-party cloud servers. In this thesis, a few novel techniques are introduced to address these problems. We begin by protecting the privacy and authenticity of transmitted readings without disrupting the direct operations. We propose two steganography techniques that rely on different mathematical security models. The results look promising - security: only the approved party who has the required security tokens can retrieve the hidden secret, and distortion effect with the difference between the original and protected readings that are almost at zero. This means the streams can be used in their protected form at intermediate hops or third party servers. We then improved the integrity of the transmitted protected streams which are prone to intentional or unintentional noise - we proposed a secure error detection and correction based stenographic technique. This allows legitimate recipients to (1) detect and recover any noise loss from the hidden sensitive information without privacy disclosure, and (2) remedy the received protected readings by using the corrected version of the secret hidden data. It is evident from the experiments that our technique has robust recovery capabilities (i.e. Root Mean Square (RMS)
Degree Doctor of Philosophy (PhD)
Institution RMIT University
School, Department or Centre Science
Subjects Computer System Security
Coding and Information Theory
Keyword(s) IoT
Privacy
Authenticity
Sensor Streams
Smart Meter
Lossless Compression
Recompression
Versions
Version Filter Type
Access Statistics: 110 Abstract Views, 115 File Downloads  -  Detailed Statistics
Created: Wed, 15 Nov 2017, 12:26:30 EST by Adam Rivett
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us