Use this URL to cite or link to this record in EThOS:
Title: Reinforcement learning routing algorithm in wireless mesh IoT networks
Author: Liu, Yu
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2020
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Internet of Things (IoT) has become a central part of our connected world. Apart from the devices in our home, many IoT devices are located in remote areas supporting all kinds of industrial, agricultural and scientific applications. Networks providing connectivity that cover in the scale of kilometre squares are crucial for these remote deployments. The extensively used star topology is not perfect for the rural environment as the coverage is limited by the placement of the central hub which also contributes to be a single point of failure. Mesh networks are clearly more appealing in this regard, but scalability has always been an issue for mesh networks, especially in terms of routing. Energy provisioning can also be challenging in the remote IoT deployments, as the devices can be left in isolated fields for a long period of time. In this thesis, we addressed the routing problem of mesh-based remote sensor IoT networks by introducing a distributive energy-aware reinforcement learning (RL) based routing algorithm. The proposed algorithm makes routing decisions by holistically considering the energy consumption of the network. This aims to maximise the durability of the entire network while preserving usability. Through the comparisons of simulated results in the failure rate, energy efficiency and carrier band usage rate of the networks supported by the proposed RL algorithm and the other applicable algorithms in the long-range remote IoT networks, we identified the strength of the RL routing algorithm for the remote sensor networks. This thesis also presents a detailed analysis of the RL routing algorithm progressively to demonstrate the effectiveness of the algorithm.
Supervisor: Tong, K. Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available