Cellular offloading in device-to-device communication is a challenging optimisation problem in which the improved allocation of radio resources can increase spectral efficiency, energy efficiency, throughout and reduce latency. The academic community have explored different optimisation methods on these problems and initial results are encouraging. However, there exists significant friction in the lack of a simple, configurable, open-source framework for cellular offload research. Prior research utilises a variety of network simulators and system models, making it difficult to compare results. In this paper we present GymD2D, a framework for experimentation with physical layer resource allocation problems in device-to-device communication. GymD2D allows users to simulate a variety of cellular offload scenarios and to extend its behaviour to meet their research needs. GymD2D provides researchers an evaluation platform to compare, share and build upon previous research. We evaluated GymD2D with state-of-the-art deep reinforcement learning and demonstrate these algorithms provide significant efficiency improvements.

Network virtualization is an inherent component of future internet architectures. Network resources are virtualized from the underlying substrate and elastically provisioned and offered to customers on-demand. Optimal allocation of network resources in terms of utilization, quality of service, and energy consumption has been a challenge. Existing solutions consider congestion control in a single-objective virtual network embedding (VNE) problem. This paper defines a multiple-objective VNE problem called the congestion-aware, energy-aware VNE (CEVNE). The aim is to seek a solution that saves cost, saves energy and avoids network congestion simultaneously. CEVNE modelling techniques and solution approaches apply both the weighting method and the constraint method to search for pareto-optimal solutions that produce the best compromised solutions for all three objectives. Solving VNE problem is, however, NP-hard. A heuristic solution is proposed involving a two-stage coordinated CEVNE. The node-mapping algorithm searches for the sub-optimal solutions for three objectives. The link mapping process is an SDN-based heuristic algorithm that deploys a path service and a resource monitoring application on an SDN controller. The solution is realized using SDN, Segment Routing, and open network operating system platform (ONOS) technologies. The energy minimization is implemented with a registry that keeps track of active nodes and sets inactive nodes to sleep mode. The evaluation results showed that the multiple-objective CEVNE approach is feasible and achieves its goals of optimizing the resource allocation, improving the runtime, saving the energy consumption and controlling the network congestion.

The Internet of Things (IoT) epitomizes the upcoming eminent transition in the world's economy and human lifestyle where people and various objects are correlated within networks. Data Aggregation is a technique which can be used to mitigate Big Data challenges within IoT. This paper provides an overview of various approaches for aggregation of data in IoT infrastructure. A new class of reliable Data Aggregation algorithm is discussed as well. This new class of algorithm uses a consensus based aggregation with fault tolerance methodology in Fog Computing. The new approach allows promoting adaptive behavior and more efficient delivery of the aggregation outcomes to the ascendant node(s). The proposed method is fault tolerant and deals with nodes reliability issues.

This research proposes a reliable machine learning based computational solution for human detection. The proposed model is specifically applicable for illumination-variant natural scenes in big data video frames. In order to solve the illumination variation problem, a new feature set is formed by extracting features using histogram of gradients (HoG) and linear phase quantization (LPQ) techniques, which are combined to form a single feature-set to describe features in illumination variant natural scenes. Pre-processing is applied to reduce the search space and improve results, and as the humans are in constant motion in the frames, a search space pruning algorithm is applied to reduce the search space and improve detection accuracy. Non-maximal suppression is also applied for improved performance. A Support Vector Machine (SVM) based classifier is used for fast and accurate detection. Most of the current state-of-the-art detectors face numerous problems including false, missed, and inaccurate detections. The proposed detector model shows good performance, which was validated using relevant UCF and CDW test data-sets. In order to compare the performance of the proposed methodology with the state-of-the-art detectors, some selected detected frames were chosen considering their Receiver Operating Characteristic (ROC) curves. These curves are plotted to compare and evaluate the results based on miss rates and true positives rates. The results show the proposed model achieves best results.

The consequence of road accidents that involves a motorcycle is far more fatal for the rider than the other drivers. Yet, there has not been an effective vehicle alert system that can eliminate these avoidable motorcycle accidents caused by other drivers where they fail to notice the motorcycles. One of the major flaws with the existing vehicle alert systems is that it should not treat motorcycles as same as other vehicles as they take much longer time to brake than a cars do. Therefore, this project aimed to find an effective method to identify motorcycles and alert the other drivers when motorcyclists are around them in 20-meter radius. After extensive literature review, the best method to solve the problem is to use road side infrastructure based Internet of Things (IOT) that divides the network into a set of clusters. In this method to identify a vehicle, it is identifying the driver and the rider from their smartphone application that beacons custom, unique Media Access Control (MAC) addresses via Bluetooth or Wi-Fi to the IOT probes. The probe differentiates the users, registers them when they arrive into the network, alerts the driver about motorcycles around them and removes them from the database when they move to other cluster. The whole scenario is simulated using the OMNET++ simulator and INET framework to demonstrate how the methodology works. If the concept is implemented in real-life, many valuable lives of motorists will be much safer on the road.

In this stage of Data Preservation the challenge is how to keep the attributes of the data and how to preserve the originality. It is like to keep the living part of the data. It is how the concepts of Heritage have sense. Heritage is the concrete data, it gives the interconnection to other aspects of the reality. Nowadays the physical value and the aspects of items complete the relevance of information. The relation between Preservation and Digital patterns of Heritage is well related because of the two aspects to consider: Accessibility and Context.

In today's life, images play a significant role in many Big Data application fields for various purposes. Image processing has to face the huge challenges because of images created in a digital format which leads to huge data volumes. Using Joint Photographic Experts Group 2000 (JPEG2000) compression techniques to meet the diverse type of real-time applications. Lossless compression JPEG2000 and others are used to minimize the expenditure of possessions such as hard disk space and transmission bandwidth. This experimental work shows an improved lossless color image compression that uses a wavelet based Human Visual System.This Reversible Color Compression Transform method (RCCT) produces an efficient algorithm to compress the image without loss of information. JPEG2000 as a lossless mode is utilized for bit-preserving and to refer globally for encoding and decoding processes. The Reversible Color Transform (RCT) is used in JPEG 2000 using wavelets which provide a mathematical way to encode the information in such a way that it is layered according to the level of detail by using HVS attributes in the stage of quantization. In this research, the goal of lossless image compression is to decrease the number of bits required to demand computing resources such as store and transmit images without any loss of information.

Link quality protocols employ link quality estimators to collect statistics on the wireless link either independently or cooperatively among the sensor nodes. Furthermore, link quality routing protocols for wireless sensor networks may modify an estimator to meet their needs. Link quality estimators are vulnerable against malicious attacks that can exploit them. A malicious node may share false information with its neighboring sensor nodes to affect the computations of their estimation. Consequently, malicious node may behave maliciously such that its neighbors gather incorrect statistics about their wireless links. This paper aims to detect malicious nodes that manipulate the link quality estimator of the routing protocol. In order to accomplish this task, MINTROUTE and CTP routing protocols are selected and updated with intrusion detection schemes (IDSs) for further investigations with other factors. It is proved that these two routing protocols under scrutiny possess inherent susceptibilities, that are capable of interrupting the link quality calculations. Malicious nodes that abuse such vulnerabilities can be registered through operational detection mechanisms. The overall performance of the new LQR protocol with IDSs features is experimented, validated and represented via the detection rates and false alarm rates.

Freshwater monitoring is becoming an essential activity due to limited availability of drinking water and an increasing presence of various pollutants. Tons of toxic waste added to water sources everyday contributes to the decrease in the planet's biodiversity and even an extinction of many species of animals and marine life. Many millions of birds perish each year due to waterway pollution. New technologies such as the Internet of Things (IoT), Wireless Sensor Networks and computer vision allow us to monitor fresh water sources in a continuous mode. To minimize the effects of pollution, various monitoring activities can be planned and executed for very large areas and geographical regions. This work presents a system architecture for the IoT-based multimodal data acquisition and analysis system. The idea is to deploy sensor clusters in various locations of a waterway to create a network of sensing and measuring smart devices. Every cluster of such devices can be perceived as a `thing'. Such a `thing' or a node has camera sensing modalities for a macro level pollution detection with analog sensors to measure microlevel water parameters. Our solution involves a low power microprocessor devices provisioned to capture raw data, extract features from the raw data and then transmit these data to the Cloud for further analysis and reporting. A 5G mobile network communication is used for data transmission. The Cloud server runs a software framework that supports a sophisticated analysis and trending of various environmental parameters such as surface density of water, salinity, temperature, etc. The proposed software framework has a set of computational algorithms to process features supplied by each node. These algorithms can classify features into various classes like floating objects, water salinity level, etc. An experiment to simulate the `IoT' data acquisition is conducted to validate the proposed solution. Based on a case study, this solution can be used in a real-life scenario representing as a feasible and viable solution to track pollution in urban waterways.

Artificial Endocrine Systems is the study of nature's highly evolved and advanced resource management system that sustains life. The basic units of the endocrine system are the chemical messengers, otherwise known as hormones, which in enough numbers produce cascading effects in the host organism that promotes growth, maintains metabolism and much more. This behaviour is in effect a living metaphor indicative of the high throughput, constant flux and stochastic interactions that exist in complex topologies such as large scale IoT infrastructures.

Identity theft is a serious crime where the criminals steal user's identity to access their bank accounts, personal hospital records and etc. The usage of cloud storage has allowed the hackers to attack the victims easily by accessing the victim's data stored in a remote location. Cloud storage may be convenient to allow users to store any data files and access those files from different devices and in different locations around the world. However, it can still pose security threats as with a lot of important data stored in one location is very appealing to cunning hackers. The solution of using image steganography application to hide important data or documents under an image before uploading them to the cloud storage will avoid the risks of attracting hackers or outsiders to discover the importance of the data hidden. The application uses one of simplest image steganography approach 'LSB algorithm' to achieve best quality of encoded image/stego-image. Besides, the application can be deployed via offline or online depending whether they want to upload the encoded image/stego-image to the cloud storage.

With the advent of smartphones, many people have the ability to take a photo and upload it to the internet whenever they want. Most smartphones will include basic metadata in the image file; however, this is usually limited to information about the camera itself, and not the content of the photo. This data is also easy to access, as there are many websites and applications that will extract the metadata from an image, thus it is unsuitable for personal information.While there are many applications out there which will allow a user to take and upload photos for their friends or the public to view, such as Facebook or Instagram, these applications rely on information provided along side the image for privacy settings, commenting, and image searching. Other search engines would have difficulty categorising the pictures they find on various sites, and if the image is saved and re-uploaded somewhere else, the information about them would likely be lost.Steganography provides a feasible and viable solution to these problems. By embedding information directly into the image, the information about the image will not be lost, if it is separated from its original source. The aim of the capstone is to demonstrate this aspect via an image sharing app that allows users to exchange messages and personalised information that is embedded in the image such that it is inaccessible without knowing their keys, and metadata that can be used by search engines to categorise the image and its contents. The developed application will provide a platform for users to create an account, share images, embed information within the image, and search for images other users have uploaded.

this paper introduces a new generation an error modelling of wireless indoor localization. A newly designed algorithm modelling by using Kalman Filter is developed to optimize the wireless indoor localization system based on iBeacon sensors. It is designed to detect and minimize the errors. Matlab simulations, implementation, and validation are presented in the paper.

Out of the many computing and software oriented models that are being adopted by Computer Networking, Fog Computing has captured quite a wide audience in Research and Industry. There is a lot of confusion on its precise definition, position, role and application. The Internet of Things (IOT), todays' digitized intelligent connectivity domain, demands real time response in many applications and services. This renders Fog Computing a suitable platform for achieving goals of autonomy and efficiency. This paper is a justification of the concepts, interest, approaches, and practices of Fog Computing. It describes the need for adopting this new model and investigate its prime features by elucidating the scenarios for implementing it, thereby outlining its significance in the IoT world.

In this paper, multiple detection engines with multi-layered intrusion detection mechanisms are proposed for enhancing computer security. The principle is to coordinate the results from each single-engine intrusion alert system, which seamlessly integrates with a multiple layered distributed service-oriented structure. An improved hidden Markov model (HMM) is created for the detection engine which is capable of the immunology-based self/nonself discrimination. The classifications of normal and abnormal behaviours of system calls are further examined by an advanced fuzzy-based inference process tuned by HPSOWM. Considering a real benchmark dataset from the public domain, our experimental results show that the proposed scheme can greatly shorten the training time of HMM and significantly reduce the false positive rate. The proposed HPSOWM works especially well for the efficient classification of unknown behaviors and malicious attacks.

In Cloud computing, due to non-locale and a multi-tenancy of services and resources, there is a need for sophisticated methods for the metering of efficiency, availability and utilisation of services and resources. Setup and management of computing infrastructure, that is in a state of constant flux, related uncertainties of the state of individual elements as well as levels of their usage per application or user on an hourly, daily, weekly, monthly, and yearly basis - pose serious challenges. This paper describes how the Anisotropic Network concept, with its mathematical apparatus, can be adopted to model, monitor and manage usage of the cloud computing resources and services seen as an elastic network of interacting elements that are in a constant motion. Various aspects of service utilisation prediction and anisotropic resource movements related to the novel Cloud Computing Elastic Network (C2EN) model will be discussed in the context of a mathematical model and experimental results of simulations.

In the development of large ad-hoc Wireless Sensor and Actuator Agent Networks (SANETS), a multitude of disparate problems are faced. In order for these networks to function, software must be able to effectively manage: unreliable dynamic distributed communication, the power constraints of un-wired devices, failure of hardware devices in hostile environments and the remote allocation of distributed processing tasks throughout the network. The solutions to these problems must be solved in a highly scalable manner. The paper describes the process of analysis of the requirements and presents a design of a service-oriented software infrastructure (middleware) solution for scalable ad-hoc networks, in a context of a system made of mobile sensors and actuators.

In the classical logic the inferential process is based on the modus ponens rule for which when the rule that connects the premise with the consequence is true and the premise is true, we know that the consequence is also true. Now, in the fuzzy inferential process the premise and the consequence are not logic propositions that assume the logic values true or false but are fuzzy sets that can assume values between zero and one. The classical logic rule is replaced with the fuzzy rule R. In this paper, we demonstrate that the fuzzy reasoning network is isomorphic to the Morphotronic network that includes the SVD technique. We also show that Morphotronic network is capable to simulate complex fuzzy reasoning or fuzzy inference in a completely new way.