The analytic hierarchy process multi-metric objective function for RPL

Alayed, Walaa Mohammad I. (2020) The analytic hierarchy process multi-metric objective function for RPL. PhD thesis, University of Glasgow.

Abstract

IPv6 Routing Protocol for Low Power and Lossy Networks (RPL), is based on building an acyclic graph where an Objective Function (OF) is responsible for selecting the preferred parent during Destination Oriented Directed Acyclic Graph (DODAG) construction. The overall aim of this research is to create a multi-metric objective function that can work in different application scenarios and outperform single metric and existing multi-metric functions in some situations while highlighting the situations in which it is not the best option.

The first part of this thesis evaluates the impact of the two standard single-metric OFs (MRHOF-ETX and OF0) on RPL performance at network-wide and node levels in terms of average power consumption, packet loss ratio, average end-to-end latency, and exchanged control packets. Results showed that RPL performance using MRHOF-ETX is better in terms of the packet loss ratio and average end-to-end latency compared to OF0. However, the opposite is the case for the average power consumption, where it is noticeable that OF0 consumes less power on average and is not affected by the network size, unlike MRHOF-ETX where the power consumption increases with the number of nodes. Single metric objective functions perform well according to some measures but poorly according to others. This suggests that perhaps multi-metric OFs can successfully combine the best features of several single-metric components.

The second part of this thesis, therefore, presents a new multi-metric OF based on the Analytic Hierarchy Process (AHP) decision-making algorithm. The AHP Objective Function (AHP-OF), combines a set of routing metrics to provide the best routing decision for RPL to satisfy the requirements of common IoT applications with random node placement, such as typical home automation application requirements. AHP-OF performance was evaluated in comparison to other OFs using discrete event simulation. The study shows a clear benefit of the proposed OF in terms of balancing the performance load among the nodes within the network.

The third part of this thesis evaluates the performance of AHP-OF in IoT environments where nodes are arranged in regular layouts with the DODAG root located at the edge, as might be encountered in some industrial scenarios. Here load balancing is almost impossible and poor performance is inevitable on some nodes. Even though AHP-OF may not be the best option in such cases, results established a noticeable advantage, that the bad impact of certain routing metrics on a single metric OF can be limited when integrated into AHP-OF.

In summary, this thesis analyses the multi-metric AHP-OF in various IoT application settings and demonstrates a clear benefit when compared to other single-metric OFs and previous multi-metric efforts.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Keywords: AHP-OF, RPL, OF, Objective function, IPv6, routing, 6LoWPAN.
• Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
• Colleges/Schools: College of Science and Engineering > School of Computing Science
• Supervisor's Name: MacKenzie, Dr. Lewis
• Date of Award: 2020
• Embargo Date: 3 June 2023
• Depositing User: Ms Walaa Alayed
• Unique ID: glathesis:2020-81416
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 10 Jun 2020 06:47
• Last Modified: 10 Jun 2020 06:47
• URI: https://theses.gla.ac.uk/id/eprint/81416