5G is the next cellular generation and is expected to quench the growing thirst for taxing data rates and to enable the Internet of Things. Focused research and standardization work have been addressing the corresponding challenges from the radio perspective while employing advanced features, such as network densification, massive multiple-input-multiple-output antennae, coordinated multi-point processing, inter-cell interference mitigation techniques, carrier aggregation, and new spectrum exploration. Nevertheless, a new bottleneck has emerged: the backhaul. The ultra-dense and heavy traffic cells should be connected to the core network through the backhaul, often with extreme requirements in terms of capacity, latency, availability, energy, and cost efficiency. This pioneering survey explains the 5G backhaul paradigm, presents a critical analysis of legacy, cutting-edge solutions, and new trends in backhauling, and proposes a novel consolidated 5G backhaul framework. A new joint radio access and backhaul perspective is proposed for the evaluation of backhaul technologies which reinforces the belief that no single solution can solve the holistic 5G backhaul problem. This paper also reveals hidden advantages and shortcomings of backhaul solutions, which are not evident when backhaul technologies are inspected as an independent part of the 5G network. This survey is key in identifying essential catalysts that are believed to jointly pave the way to solving the beyond-2020 backhauling challenge. Lessons learned, unsolved challenges, and a new consolidated 5G backhaul vision are thus presented.

The Internet of Things (IoT) is revolutionising how energy is delivered from energy producers and used throughout residential households. Optimising the residential energy consumption is a crucial step toward having greener and sustainable energy production. Such optimisation requires a household-centric energy management system as opposed to a one-rule-fits all approach. In this paper, we propose a data-driven multi-layer digital twin of the energy system that aims to mirror households’ actual energy consumption in the form of a household digital twin (HDT). When linked to the energy production digital twin (EDT), HDT empowers the household-centric energy optimisation model to achieve the desired efficiency in energy use. The model intends to improve the efficiency of energy production by flattening the daily energy demand levels. This is done by collaboratively reorganising the energy consumption patterns of residential homes to avoid peak demands whilst accommodating the resident needs and reducing their energy costs. Indeed, our system incorporates the first HDT model to gauge the impact of various modifications on the household energy bill and, subsequently, on energy production. The proposed energy system is applied to a real-world IoT dataset that spans over two years and covers seventeen households. Our conducted experiments show that the model effectively flattened the collective energy demand by $20.9\%$ on synthetic data and $20.4\%$ on a real dataset. At the same time, the average energy cost per household was reduced by 10.7% for the synthetic data and 17.7% for the real dataset.

The Internet of Things (IoT) paradigm is revolutionising the world of manufacturing into what is known as Smart Manufacturing or Industry 4.0. The main pillar in smart manufacturing looks at harnessing IoT data and leveraging machine learning (ML) to automate the prediction of faults, thus cutting maintenance time and cost and improving the product quality. However, faults in real industries are overwhelmingly outweighed by instances of good performance (faultless samples); this bias is reflected in the data captured by IoT devices. Imbalanced data limits the success of ML in predicting faults, thus presents a significant hindrance in the progress of smart manufacturing. Although various techniques have been proposed to tackle this challenge in general, this work is the first to present a framework for evaluating the effectiveness of these remedies in the context of manufacturing. We present a comprehensive comparative analysis in which we apply our proposed framework to benchmark the performance of different combinations of algorithm components using a real-world manufacturing dataset. We draw key insights into the effectiveness of each component and inter-relatedness between the dataset, the application context, and the design of the ML algorithm.

Intelligent transport systems (ITS) are pivotal in the development of sustainable and green urban living. ITS is data-driven and enabled by the profusion of sensors ranging from pneumatic tubes to smart cameras which are used to detect and categorise passing vehicles. Simple sensors, such as pneumatic tubes, are successfully deployed for counting passing vehicles but are not useful for vehicle tracking or re-identification. Smart cameras, on the other hand, collect comprehensive information but suffer from occlusion, patchy coverage, and compromised vision in adverse weather and visibility. This work explores a novel ITS data source based on optical fibre which acts as uninterrupted length of virtual sensors using a distributed acoustic sensor (DAS) system. Based on real DAS data collected in the field, we first present a study of latent DAS features that uniquely identify a given vehicle, otherwise referred to as vehicle signature. We formulate a classification problem that examines incoming DAS data to extract vehicle signatures and identify the different types of vehicle. To this end, we implement different classification methods and present a comparative performance analysis that reveals novel insights into the potential role of DAS in ITS applications. This work is a pilot study of DAS for vehicle classification that is driven by real-DAS data and validated by promising results where a vehicle type is correctly identified with 94% accuracy and the size of a vehicle with 95% accuracy.

Wireless technology is the strongest contender for catering for the 5G backhaul (BH) stipulated performance, where optical fiber is unavailable. In the presence of ultra-dense networks, such occurrences are exponentially increasing, and different wireless technologies are investigated for this application. We present the first BH-specific wireless link performance modeling that considers its inherent line-of-sight nature, together with an appropriate representation of the network topology using stochastic geometry. To this end, novel tractable models are obtained to capture the performance of wireless BH links. These are integrated into a multi-hop hybrid BH performance modeling framework and are applied in the analysis of a BH-aware user association optimization problem.

5G definition and standardization projects are well underway, and governing characteristics and major challenges have been identified. A critical network element impacting the potential performance of 5G networks is the backhaul, which is expected to expand in length and breadth to cater to the exponential growth of small cells while offering high throughput in the order of gigabit per second and less than 1 ms latency with high resilience and energy efficiency. Such performance may only be possible with direct optical fiber connections that are often not available country-wide and are cumbersome and expensive to deploy. On the other hand, a prime 5G characteristic is diversity, which describes the radio access network, the backhaul, and also the types of user applications and devices. Thus, we propose a novel, distributed, self-optimized, end-to-end user-cell-backhaul association scheme that intelligently associates users with candidate cells based on corresponding dynamic radio and backhaul conditions while abiding by users' requirements. Radio cells broadcast multiple bias factors, each reflecting a dynamic performance indicator (DPI) of the end-to-end network performance such as capacity, latency, resilience, energy consumption, and so on. A given user would employ these factors to derive a user-centric cell ranking that motivates it to select the cell with radio and backhaul performance that conforms to the user requirements. Reinforcement learning is used at the radio cells to optimise the bias factors for each DPI in a way that maximise the system throughput while minimising the gap between the users' achievable and required end-to-end quality of experience (QoE). Preliminary results show considerable improvement in users' QoE and cumulative system throughput when compared with the state-of-the-art user-cell association schemes.