Disputation Cenk Gündogan
Cenk defends his PhD thesis on "Information-centric Networking for the Constrained Internet of Things"
- https://www.inet.haw-hamburg.de/events/community/disputation-cenk-gundogan
- Disputation Cenk Gündogan
- 2022-07-19T16:00:00+02:00
- 2022-07-19T18:30:00+02:00
- Cenk defends his PhD thesis on "Information-centric Networking for the Constrained Internet of Things"
Jul 19, 2022 from 04:00 PM to 06:30 PM (Europe/Berlin / UTC200)
The deployment of Information-Centric Networking (ICN) on embedded wireless devices was first considered a decade ago. With the advent of the Internet of Things (IoT), early experiments confirmed benefits for constrained multi-hop networks from operating Named-Data Networking (NDN) as a network layer directly on top of data links. Since then, a large body of work has proposed and evaluated ICN in the IoT context. Essential findings show that hop-wise forwarding with caching and a leaner network stack improve adaptability to lossy regimes and network performance over IP-style communication with end-to-end transport services. While a native deployment of the information-centric content replication already promotes a reliable and energy-efficient IoT edge, additional protocol extensions further reduce memory demands, improve the utilization of wireless network links, and decrease the overall power consumption.
Part I of this manuscript aims to develop an information-centric protocol suite that integrates with off-the-shelf IoT hardware and software platforms. A new convergence layer follows the design considerations of IPv6 over Low-Power Wireless Personal Area Networks (6LoWPAN), and adapts ICN packets to the small-sized Maximum Transmission Units (MTUs) of low-power links by providing (i) a stateless and stateful header compression scheme, (ii) link fragmentation, and (iii) protocol framing similar to EtherTypes. In addition to this link convergence, which operates below the actual ICN layer, a lightweight Quality of Service (QoS) scheme on the network layer improves the utilization of limited network buffers. It enables a coordinated resource management to reduce network latency and to prevent queue starvation for prioritized traffic flows. Since an intermittent connectivity and device mobility are common in wireless regimes of low reliability, this protocol suite further adds a lightweight publish–subscribe system to bolster the information-centric IoT against network disruptions, and to improve routing agility on connectivity loss. Part II examines the Internet perspective of native ICN IoT networks and then describes the construction of a data-centric Web of Things (WoT). This deployment option is based on standard protocol elements of the Constrained Application Protocol (CoAP), and replicates the information-centric principles (i) stateful forwarding, (ii) hop-wise caching, and (iii) content object security. Real protocol implementations and testbed assessments on actual IoT hardware show that the data-centric WoT adheres to the performance expectations of pure ICN deployments, while retaining full compatibility with Internet services. Additionally, this deployment option leads insights and techniques that emerged from ICN research into a promising, realistic deployment trail for the growing IoT.