Please use this identifier to cite or link to this item:
Title: Network performance evaluation for M2M WSN and SDN based on IOT applications
Authors: Twayej, Wasan Adnan
Advisors: Al-Raweshidy, H
Abbod, M
Keywords: MLCMS;SDN-WISE;SOCHSA;DPSO;Adaptive sleep mode
Issue Date: 2018
Publisher: Brunel University London
Abstract: This thesis introduces different mechanisms for energy efficiency in Wireless Sensor Networks (WSNs) along with maintaining high levels of Network Performance (N.P) with reduced complexity. Firstly, a Machine-to-Machine (M2M) WSN is arranged hierarchically in a fixed infrastructure to support a routing protocol for energy-efficient data transmission among terminal nodes and sink nodes via cluster heads (CHs). A Multi-Level Clustering Multiple Sinks (MLCMS) routing protocol with the IPv6 protocol over Low Wireless Personal Area Networks (6LoWPAN) is proposed to prolong network lifetime. The simulation results show 93% and 147% enhancement in energy efficiency and system lifespan compared to M-LEACH and LEACH, respectively. By utilising 6LoWPAN in the proposed system, the number of packets delivered increases by 7%, with higher accessibility to the M2M nodes and a substantial extension of the network is enabled. Secondly, an adaptive sleep mode with MLCMS for an efficient lifetime of M2M WSN is introduced. The time period of the active and asleep modes for the CHs has been considered according to a mathematical function. The evaluations of the proposed scheme show that the lifetime of the system is doubled and the end-to-end delay is reduced by half. Thirdly, enhanced N.P is achieved through linear integer-based optimisation. A Self-Organising Cluster Head to Sink Algorithm (SOCHSA) is proposed, hosting Discrete Particle Swarm Optimisation (DPSO) and Genetic Algorithm (GA) as Evolutionary Algorithms (EAs) to solve the N.P optimisation problem. N.P is measured based on load fairness and average ratio residual network energy. DPSO and GA are compared with the Exhaustive Search (ES) algorithm to analyse their performances for each benchmark problem. Computational results prove that DPSO outperforms GA regarding complexity and convergence, thus it is best suited for a proactive IoT network. Both algorithms achieved optimum N.P evaluation values of 0.306287 and 0.307731 in the benchmark problems P1 and P2, respectively, for two and three sinks. The proposed mechanism satisfies different N.P requirements of M2M traffic by instant identification and dynamic rerouting to achieve optimum performance. Finally, a Power Model (PM) is essential to investigate the power efficiency of a system. Hence, a Power Consumption (PC) profile for SDN-WISE, based on IoT is developed. The outcomes of the study offer flexibility in managing the structure of an M2M system in IoT. They enable controlling the provided Network Quality of Service (NQoS), precisely by achieving physical layer throughput. In addition, it provides a schematic framework for the Application Quality of Service (AQoS), specifically, the IoT data stream payload size (from the PC point of view). It is composed of two essential parts, i.e., control signalling and data traffic PCs and the results show a 98% PC of the data plane in the total system power, whereas the control plane PC is only 2%, with a minimum Transmission Time Interval (TTI) (5 sec) and a maximum payload size of 92 Bytes.
Description: This thesis was submitted for the award of Doctor of Philosophy and was awarded by Brunel University London
Appears in Collections:Electronic and Computer Engineering
Dept of Electronic and Electrical Engineering Theses

Files in This Item:
File Description SizeFormat 
FulltextThesis.pdf3.09 MBAdobe PDFView/Open

Items in BURA are protected by copyright, with all rights reserved, unless otherwise indicated.