| Titre : |
Q-learning-based dynamic transmission channel selection for Wireless sensor networks |
| Type de document : |
document multimédia |
| Auteurs : |
Safa Benaidja, Auteur ; Bouchra Laadjal, Auteur ; Lakhdar Kamel Oulad Djedid, Directeur de thèse |
| Editeur : |
Laghouat : Université Amar Telidji - Département d'informatique |
| Année de publication : |
2026 |
| Importance : |
97 p. |
| Accompagnement : |
1 disque optique numérique (CD-ROM) |
| Note générale : |
Option : Distributed networks, systems and applications |
| Langues : |
Anglais (eng) |
| Mots-clés : |
Wireless sensor networks Q-Learning Reinforcement learning Dynamic channel selection IEEE 802.15.4 Interference management Markov decision process |
| Résumé : |
Wireless Sensor Networks (WSNs) are increasingly used in applications such as environmental monitoring, industrial automation, and smart systems, where reliable and energy-efficient communication is essential. Since WSNs operate in unlicensed frequency bands, wireless channels are highly affected by interference and dynamic channel conditions, leading to packet collisions, transmission failures, and reduced network performance. Traditional static and random channel allocation methods are unable to adapt efficiently to these varying conditions.
This thesis proposes a Q-learning-based dynamic channel selection framework for WSNs, where each sensor node acts as an independent reinforcement learning agent capable of autonomously selecting transmission channels according to observed interference conditions. The problem is modeled as a Markov Decision Process (MDP) with a three-level interference state space (Low, Medium, High) and an action space composed of eight IEEE 802.15.4-compatible channels. Channel selection is performed using an ϵ-greedy policy to balance exploration and exploitation during learning.
The proposed framework was evaluated through Python-based simulations under low, high, and dynamic interference scenarios using network sizes ranging from 5 to 30 sensor nodes. Performance evaluation was conducted using Packet Delivery Ratio (PDR), Throughput, Energy Consumption, and End-to-End Delay. Simulation results demonstrate that the proposed Q-learning approach significantly improves channel selection efficiency and consistently outperforms both random channel selection and static channel allocation methods by reducing interference effects and improving communication reliability. |
| note de thèses : |
Mémoire de master en informatique |
Q-learning-based dynamic transmission channel selection for Wireless sensor networks [document multimédia] / Safa Benaidja, Auteur ; Bouchra Laadjal, Auteur ; Lakhdar Kamel Oulad Djedid, Directeur de thèse . - Laghouat : Université Amar Telidji - Département d'informatique, 2026 . - 97 p. + 1 disque optique numérique (CD-ROM). Option : Distributed networks, systems and applications Langues : Anglais ( eng)
| Mots-clés : |
Wireless sensor networks Q-Learning Reinforcement learning Dynamic channel selection IEEE 802.15.4 Interference management Markov decision process |
| Résumé : |
Wireless Sensor Networks (WSNs) are increasingly used in applications such as environmental monitoring, industrial automation, and smart systems, where reliable and energy-efficient communication is essential. Since WSNs operate in unlicensed frequency bands, wireless channels are highly affected by interference and dynamic channel conditions, leading to packet collisions, transmission failures, and reduced network performance. Traditional static and random channel allocation methods are unable to adapt efficiently to these varying conditions.
This thesis proposes a Q-learning-based dynamic channel selection framework for WSNs, where each sensor node acts as an independent reinforcement learning agent capable of autonomously selecting transmission channels according to observed interference conditions. The problem is modeled as a Markov Decision Process (MDP) with a three-level interference state space (Low, Medium, High) and an action space composed of eight IEEE 802.15.4-compatible channels. Channel selection is performed using an ϵ-greedy policy to balance exploration and exploitation during learning.
The proposed framework was evaluated through Python-based simulations under low, high, and dynamic interference scenarios using network sizes ranging from 5 to 30 sensor nodes. Performance evaluation was conducted using Packet Delivery Ratio (PDR), Throughput, Energy Consumption, and End-to-End Delay. Simulation results demonstrate that the proposed Q-learning approach significantly improves channel selection efficiency and consistently outperforms both random channel selection and static channel allocation methods by reducing interference effects and improving communication reliability. |
| note de thèses : |
Mémoire de master en informatique |
|  |
Ajouter le résultat dans votre panier Faire une suggestion Affiner la rechercheQ-learning-based dynamic transmission channel selection for Wireless sensor networks / Safa Benaidja
