Democritus University of Thrace (DUTH)
NDN- and DTN-based solutions for intermittent IoT communications
Abstract
dc:descriptionNowadays, modern infrastructure, vehicles, devices and physical objects sense, collect and communicate immense amounts of data with each other and with the Internet, shaping the so-called Internet of Things (IoT). The rapid technological evolution and broad applicability of IoT, paved the way for its widespread deployment in network environments with distinct characteristics and connectivity issues. One of the most critical issues for the well-functioning of modern IoT systems concerns the reliable and efficient support of intermittent communication between IoT devices, which is mainly caused due to nodes' mobility, poor coverage, limited connectivity or even due to high signal interference. Since the traditional TCP/IP stack faces difficulties with intermittently-connected IoT networks, because it is designed for the end-to-end interconnection of continuously connected devices, alternative networking architectures such as Named Data Networking (NDN) and Delay-/Disruption-Tolerant Networking (DTN) are actively investigated. NDN is one of the most well-known Information-Centric Networking (ICN) architectures, which is based on the content naming and distribution in a location-independent manner. Thus, NDN largely enhances content availability and data retrieval efficiency, however is unable to support highly volatile IoT environments, due to its use of symmetric paths. Contrarily, DTN architecture adopts the store-carry-and-forward approach, which allows nodes to stay interconnected in scenarios of prolonged delays and disruptions, by leveraging each available intermediate node, path and network. Although DTN improves the reliability of intermittently connected IoT networks, its host-centric communication model impedes the support of data-centric IoT communications. Taking stock of the distinct benefits of each approach, this thesis proposes NDN- and DTN-based combined and adaptive solutions to boost network performance and reliability, support mobility, enhance interoperability and bring communication flexibility to intermittently-connected IoT environments. In the first part of this thesis, we propose combined NDN- and DTN-based architectures to support mobility, enable interoperability and enhance communication versatility, in the context of disruptive IoT networks. Initially, we focus on enabling the flawless operation of NDN-based devices over highly unstable and mobile IoT environments, by proposing an integrated NDN-DTN scheme, called NoD, as well as cross-layer mechanisms between the NDN and DTN layers. Then, we address interoperability and client mobility issues of the TCP/IP-based MQTT protocol, by combining the recently introduced MQTT 5.0 standard with the DTN architecture. Interoperability-wise, we propose a combined edge architecture that enables the interconnection of continuously-connected MQTT 5.0 and partially-connected DTN devices, through its gateway-based operations. With respect to MQTT client mobility support, we introduce a proxy-based architecture, called MD-IoT, to ensure the seamless and reliable communication among mobile MQTT 5.0 publishers and subscribers, through the transparent integration of the DTN layer. Afterwards, we propose a universal multi-stack architecture, called MeND, that combines MQTT 5.0, NDN and DTN to support multiple IoT operations as well as flexible data-centric communication patterns over both continuously- and partially-connected IoT networks. We thoroughly assessed the performance, reliability and effectiveness of the proposed combined architectures, through extensive experimental campaigns in real-world and emulated opportunistic and mobile networks, stationary fully-connected and lossy environments and wireless smart city networks. The evaluation results indeed verify that the integrated NoD scheme and its cross-layer extensions, largely improve the reliability and performance of diverse disruption-prone IoT scenarios and, especially in opportunistic and highly disruptive environments, can outperform the effectiveness of NDN and DTN. Concerning our joint MQTT 5.0 and DTN endeavors, results indicate that: (i) our edge architecture can seamlessly operate under various disruption periods, allowing MQTT 5.0 and DTN devices to stay interconnected even in extreme conditions, and (ii) MD-IoT enables the efficient, reliable and transparent communication of mobile MQTT 5.0 publishers and subscribers. Also, our experimental findings validate the robust and efficient operation of the universal MeND architecture and highlight its multifunctional capabilities and communication versatility to support fully- and partially-connected IoT devices with diverse computational resources. In the second part of this dissertation, inspired by the distinct benefits and divergent performance of NDN, DTN and NoD, we propose adaptive Software-Defined Networking (SDN) based systems that leverage interchangeably and on-demand each approach to improve the performance and reliability of smart city networks with dynamically changing network conditions. First, we introduce a threshold-based adaptive system that collects urban network monitoring data, detects the connectivity state and the reliable paths of the network and accordingly deploys the appropriate protocol stack to the smart city network nodes. In the next phase, we focused on enabling protocol-adaptive strategies over wireless smart city links that experience periods of reliable and unreliable connectivity, and grafted SDN control features, link-based hybrid NDN, DTN and NoD strategies along with link reliability detection mechanisms in a single smart city platform, called REWIRE. Finally, we consolidate and extend our prior adaptive systems, into an adaptive SDN-based multi-protocol system, called ALTER, that incorporates mixed decision-making capabilities, experiment automation and network emulation facilities as well as link- and network-based adaptive protocol deployment and configuration operations. The evaluation results of the proposed adaptive systems, which are deduced from our extensive experimental campaign in real-world wireless smart city networks with dynamic network conditions, indicate that the interchangeable employment of NDN, DTN and NoD largely boost the performance and reliability of wireless urban networks, even by adopting link-based or cumulative network-based approaches. Specifically, our experimental findings underscore the clear-cut advantage of our proposed adaptive systems, compared to the individual NDN and NoD deployments under various dynamic network conditions and single-link and multi-hop wireless urban networks.
Degree
thesis:*- Grantor dc:publisher
- Democritus University of Thrace (DUTH)
- Year dc:date
- 2024
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Demiroglou, Vasileios
- Δεμίρογλου, Βασίλειος
Subjects
dc:subject × 20- Πληροφοριοκεντρικά δίκτυα
- Δίκτυα ανεκτικά σε καθυστερήσεις και διακοπές
- Διαδίκτυο των πραγμάτων
- Έξυπνες πόλεις
- Information-centric networking
- Delay-/Disruption- tolerant networking
- Internet of things
- Smart cities
- Επιστήμη Ηλεκτρονικών Υπολογιστών και Πληροφορική
- Φυσικές Επιστήμες
- Δίκτυα ηλεκτρονικών υπολογιστών και Επικοινωνίες
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Επιστήμες Μηχανικού και Τεχνολογία
- Μηχανική και συστήματα επικοινωνιών, Τηλεπικοινωνίες
- Computer and Information Sciences
- Natural Sciences
- Computer Networks and Communications
- Electrical Engineering, Electronic Engineering, Information Engineering
- Engineering and Technology
- Communication engineering and systems, Telecommunications
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/57950