Abstract
dc:descriptionThe rapid growth of the Internet of Things (IoT) and Wireless Sensor Networks (WSNs) signifies a fundamental shift in the methods of data sensing, processing, and transmission across several application domains. The deployment of large IoT networks, particularly those utilising Low-Power Wide-Area Network (LPWAN) technologies, such as LoRaWAN, encounters significant challenges concerning energy efficiency, scalability, fairness, and reliable data transmission. These constraints are intensified by LoRaWAN’s ALOHA-based Medium Access Control (MAC) protocol, resulting in increased collision rates, inadequate spectrum utilisation, and restricted capacity for varying data loads. This dissertation investigates these challenges through a layered methodology that begins with a comprehensive theoretical analysis of IoT communication principles, architectures, and existing protocol limitations, with emphasis on the operational dynamics of LoRaWAN. Building upon this foundational understanding, the research introduces novel MAC-layer scheduling schemes, designed to enhance performance across key network metrics. The first major contribution is Fair Collision Avoidance for LoRa (FCA-LoRa), a beacon-based scheduling protocol designed to support fair collision avoidance. FCA-LoRa incorporates superframe structuring and Carrier Sense Multiple Access with Collision Avoidance (CSMA-CA) protocol for accessing the medium to improve network coordination and minimise energy-intensive retransmissions. Following this, Time-Slotted LoRa with Variable Payload Support (TS-VPLoRa) is developed as an advanced time-slotted MAC protocol capable of managing variable payload transmissions, addressing challenges related to spreading factor (SF) collisions and spectrum underutilisation in mixed-traffic environments. Beyond MAC-level enhancements, the dissertation proposes LoRA-Federated Learning (LoRAFL), an architecture that integrates Federated Learning (FL) with LoRaWAN to enable lightweight, decentralised Artificial Intelligence (AI) model distribution. The methodology includes the design and implementation of a knowledge distillation pipeline, enabling efficient model compression and transmission under strict LPWAN constraints, leveraging FCA-LoRa. LoRA-FL focuses on enabling real-world intelligent IoT applications in a privacy-preserving and resource-aware manner. All proposed solutions were rigorously validated through simulations using OMNeT++ to model realistic LoRaWAN deployments. Experimental scenarios were designed across varied node densities and payload conditions to assess throughput, fairness, latency, and energy consumption. Specific focus was given to the simulation of AI model distribution processes in LoRA-FL, where multiple gateways served as model aggregators to assess scheduling efficiency and communication overhead. The results demonstrate considerable improvements in throughput, energy efficiency, scalability, and adaptability. These findings establish a robust basis for the development of future LPWAN systems capable of integrating intelligent, decentralised processing and adaptive scheduling. This work contributes to redefining MAC-layer design for next-generation IoT deployments, while setting a foundation for AI-enhanced, real-time, and resource-efficient IoT communication infrastructures.
Degree
thesis:*- Grantor dc:publisher
- University of Western Macedonia
- Year dc:date
- 2025
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Triantafyllou, Anna
- Τριανταφύλλου, Άννα
Subjects
dc:subject × 19- Διαδικτύου των Πραγμάτων (ΔτΠ)
- Δίκτυα ευρείας κάλυψης με χαμηλή κατανάλωση ενέργειας (LPWAN)
- Χρονοπρογραμματισμός
- Δρομολόγηση
- LoRaWAN
- Επίπεδο Πολλαπλής Πρόσβασης στο μέσο (MAC)
- Αρχιτεκτονική με βάση την τεχνολογία μηχανικής μάθησης Federated Learning
- Internet of things (IoT)
- Low-Power Wide-Area Network (LPWAN)
- Scheduling
- Routing
- Medium Access Control (MAC)
- Federated Learning-based architecture
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Επιστήμες Μηχανικού και Τεχνολογία
- Μηχανική και συστήματα επικοινωνιών, Τηλεπικοινωνίες
- 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/60434