Classen, Jiska (2020)
Security and Privacy for IoT Ecosystems.
Technische Universität Darmstadt
doi: 10.25534/tuprints-00011422
Ph.D. Thesis, Primary publication
|
Text
2020-02-04_document.pdf Copyright Information: CC BY-NC-ND 4.0 International - Creative Commons, Attribution NonCommercial, NoDerivs. Download (8MB) | Preview |
Item Type: | Ph.D. Thesis | ||||
---|---|---|---|---|---|
Type of entry: | Primary publication | ||||
Title: | Security and Privacy for IoT Ecosystems | ||||
Language: | English | ||||
Referees: | Hollick, Prof. Dr. Matthias ; Patras, Ass. Prof. Paul | ||||
Date: | 20 January 2020 | ||||
Place of Publication: | Darmstadt | ||||
Date of oral examination: | 20 January 2020 | ||||
DOI: | 10.25534/tuprints-00011422 | ||||
Abstract: | Smart devices have become an integral part of our everyday life. In contrast to smartphones and laptops, Internet of Things (IoT) devices are typically managed by the vendor. They allow little or no user-driven customization. Users need to use and trust IoT devices as they are, including the ecosystems involved in the processing and sharing of personal data. Ensuring that an IoT device does not leak private data is imperative. This thesis analyzes security practices in popular IoT ecosystems across several price segments. Our results show a gap between real-world implementations and state-of-the-art security measures. The process of responsible disclosure with the vendors revealed further practical challenges. Do they want to support backward compatibility with the same app and infrastructure over multiple IoT device generations? To which extent can they trust their supply chains in rolling out keys? Mature vendors have a budget for security and are aware of its demands. Despite this goodwill, developers sometimes fail at securing the concrete implementations in those complex ecosystems. Our analysis of real-world products reveals the actual efforts made by vendors to secure their products. Our responsible disclosure processes and publications of design recommendations not only increase security in existing products but also help connected ecosystem manufacturers to develop secure products. Moreover, we enable users to take control of their connected devices with firmware binary patching. If a vendor decides to no longer offer cloud services, bootstrapping a vendor-independent ecosystem is the only way to revive bricked devices. Binary patching is not only useful in the IoT context but also opens up these devices as research platforms. We are the first to publish tools for Bluetooth firmware and lower-layer analysis and uncover a security issue in Broadcom chips affecting hundreds of millions of devices manufactured by Apple, Samsung, Google, and more. Although we informed Broadcom and customers of their technologies of the weaknesses identified, some of these devices no longer receive official updates. For these, our binary patching framework is capable of building vendor-independent patches and retrofit security. Connected device vendors depend on standards; they rarely implement lower-layer communication schemes from scratch. Standards enable communication between devices of different vendors, which is crucial in many IoT setups. Secure standards help making products secure by design and, thus, need to be analyzed as early as possible. One possibility to integrate security into a lower-layer standard is Physical-Layer Security (PLS). PLS establishes security on the Physical Layer (PHY) of wireless transmissions. With new wireless technologies emerging, physical properties change. We analyze how suitable PLS techniques are in the domain of mmWave and Visible Light Communication (VLC). Despite VLC being commonly believed to be very secure due to its limited range, we show that using VLC instead for PLS is less secure than using it with Radio Frequency (RF) communication. The work in this thesis is applied to mature products as well as upcoming standards. We consider security for the whole product life cycle to make connected devices and IoT ecosystems more secure in the long term. |
||||
Alternative Abstract: |
|
||||
URN: | urn:nbn:de:tuda-tuprints-114229 | ||||
Classification DDC: | 000 Generalities, computers, information > 004 Computer science | ||||
Divisions: | 20 Department of Computer Science > Sichere Mobile Netze | ||||
Date Deposited: | 08 May 2020 10:02 | ||||
Last Modified: | 09 Jul 2020 06:25 | ||||
URI: | https://tuprints.ulb.tu-darmstadt.de/id/eprint/11422 | ||||
PPN: | 46514103X | ||||
Export: |
View Item |