Energy Hole Mitigation and Network Lifetime Maximisation in Shape-Varying 3D IoT-based Heterogeneous Wireless Sensor Networks

The rapid expansion of IoT-based Heterogeneous Wireless Sensor Networks (HWSNs) amplifies challenges in energy management. These networks, comprising diverse devices deployed across varying terrains, are designed to efficiently collect data while ensuring complete coverage. However, achieving balanced energy consumption is critical to prevent energy holes, which occur when devices near base station deplete their energy at faster rate due to experiencing excessive transmission loads. Such imbalances lead to network instability and reduce operational lifespan. Existing energy hole mitigation techniques often fail to account for the diverse characteristics of heterogeneous devices, limiting their effectiveness and applicability. Furthermore, many of these approaches are tailored to specific applications or constrained by fixed network shapes, lacking the adaptability, scalability, and flexibility required for large-scale, dynamic IoT-based HWSNs. These limitations hinder their performance in managing the complex and diverse topologies encountered in real-world deployments.

This thesis aims to tackle these challenges by proposing novel methods to reduce energy consumption and balance it across resource-constrained heterogeneous devices. The proposed solutions focus on enhancing energy efficiency, extending network lifetime, and improving throughput, ensuring reliable data transmission from sensing nodes to the base station.

A critical focus is on deployment of the base station, is critical given the converge-caste nature of communication in HWSNs. An iterative algorithm is proposed to compute the overall network energy consumption for a set of alternative locations within a specified radius of the base station. While the iterative algorithm identifies the optimal deployment location, its computational complexity is addressed using a multi-criteria decision-making approach. The Technique for Order of Preference by Similarity to Ideal Solutions (TOPSIS) evaluates alternative locations based on their positive and negative contributions to energy consumption and balancing. This method improves network lifetime by 23.5% compared to central deployment techniques and adapts effectively to networks of varying shapes, including circular, square, cubical and spherical, in both 2D and 3D environments.

In the post-deployment phase, the distribution of heterogeneous resources across smaller network segments supports the design of an effective communication topology. Considering variations in network shape and spatial dimensions, this thesis proposes two fixed-shape segmentation schemes-cubical and spherical segmentation, which cover the majority of network geometries. These schemes leverage the distribution functions of network parameters and integrate an unequal clustering method to ensure energy-balanced routing. However, the complexity of choosing the appropriate segmentation scheme for specific applications motivates the development of a novel shape-independent, data-traffic-based segmentation approach. To overcome this challenge, a novel shape-independent, data-traffic-based segmentation scheme is introduced. This scheme provides faster and more accurate insights into network parameter distributions, enhancing the unequal clustering algorithm. The integration of shape-independent segmentation improves network lifetime by up to 18.8% and reduces energy consumption by 61.4% compared to existing clustering methods.

Finally, coordinated and dynamic energy-efficient mechanisms for cluster head and next-hop relay node selection, along with rotation strategies, further enhance energy efficiency. These methods optimise both intra-cluster and inter-cluster communication by dynamically selecting energy-efficient cluster head nodes, rotating their roles, and identifying optimal next-hop relay nodes for inter-cluster communication. The integration of appropriate rotation mechanisms ensures balanced energy consumption across the network. The proposed routing techniques improve network throughput by 57.44% and increase energy efficiency by 17.63% compared to protocols such as Rotated Low-Energy Adaptive Clustering Hierarchy (RLEACH) and Cluster Routing Protocol using Fuzzy C-Means (CRPFCM).

The methods proposed in this thesis significantly extend network lifetime and optimise energy usage in IoT-based HWSNs. They provide a scalable and adaptable framework for energy management in complex, multi-parameter, and multi-level IoT-based HWSNs,