Use this URL to cite or link to this record in EThOS:
Title: Cognitive network framework for heterogeneous wireless mesh systems
Author: Al-Saadi, Ahmed
ISNI:       0000 0004 5989 3827
Awarding Body: Cardiff University
Current Institution: Cardiff University
Date of Award: 2016
Availability of Full Text:
Access from EThOS:
Access from Institution:
Heterogeneous wireless mesh networks (WMN) provide an opportunity to secure higher network capacity, wider coverage and higher quality of service (QoS). However, heterogeneous systems are complex to configure because of the high diversity of associated devices and resources. This thesis introduces a novel cognitive network framework that allows the integration of WMNs with long-term evolution (LTE) networks so that none of the overlapped frequency bands are used. The framework consists of three novel systems: the QoS metrics management system, the heterogeneous network management system and the routing decision-making system. The novelty of the QoS metrics management system is that it introduces a new routing metric for multi-hop wireless networks by developing a new rate adaptation algorithm. This system directly addresses the interference between neighbouring nodes, which has not been addressed in previous research on rate adaptation for WMN. The results indicated that there was a significant improvement in the system throughput by as much as to 90%. The routing decision-making system introduces two novel methods to select the transmission technology in heterogeneous nodes: the cognitive heterogeneous routing (CHR) system and the semantic reasoning system. The CHR method is used to develop a novel reinforcement learning algorithm to optimise the selection of transmission technology on wireless heterogeneous nodes by learning from previous actions. The semantic reasoning method uses ontologies and fuzzy-based semantic reasoning to facilitate the dynamic addition of new network types to the heterogeneous network. The simulation results showed that the heterogeneous network outperformed the benchmark networks by up to 200% of the network throughput.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: TK Electrical engineering. Electronics Nuclear engineering