The Millimetre-wave (mmWave) frequency band offers wide available bandwidth for 5G and future wireless networks. However, due to the large propagation attenuation of the mmWave frequency, real-world environmental factors, including weather, foliage, humans, vehicles, have a significant impact on the channel path loss. In order to design and optimise 5G and future networks, a mmWave channel model considering the environmental factors is urgently required. In this project, we aim to equip a ray-tracing channel model with environmental factors in the mmWave band. The goal will be approached through 3 work packages (WPs). First, the ER will develop a parametric attenuation model by parameterising the environmental factors and integrate it with a ray-tracing simulation model. Second, the ER will acquire the channel data for each corresponding environment factor and extract parameters using machine-learning-based methods. Third, the developed channel model will be applied to evaluate and optimise mmWave wireless networks taking environmental factors into account. This project will be an extension of the ER’s previous work, which will develop him to be a mature researcher.

Network slicing is a key enabling technology for the 5th generation (5G) and beyond mobile networks. Network slicing allows the creation of multiple logical network instances on the same underlying physical network. Slices can then be formed or combined on-demand, with parameters optimized according to various service requirements so as to meet the users’ instant requests for specific mobile services. Hence, the performance of network slicing heavily depends on characterizing and predicting the spatial- temporal traffic patterns for individual services in near real-time. Research on characterization and prediction for service-level mobile traffic is still in nascence. Firstly, the traffic characteristics and predicting methods of individual services, especially the 5G services, have not been studied adequately. Secondly, traffic correlation among different services and the reasons behind it have not been well studied. Thirdly, inter-service correlations have not been well exploited in service-level mobile traffic prediction. In this project, we will address these gaps. Firstly, we will study the spatial-temporal characteristics of service-level traffic patterns at multi-scales, based on which, we will investigate the traffic predicting frameworks for individual services. Secondly, for the first time, we will investigate the traffic correlation among different services and try to discover the underlying reasons by analyzing the service usage profiles of different user groups. Finally, based on inter-service correlations, we will investigate whether we can improve service-level traffic prediction accuracy and whether we could execute prediction for diverse services according to the historical records of only a few key services. The success of the CORRELATION project will make proactive network slicing possible, which will then drive proactive network optimisation for 5G and beyond mobile networks.

The foreseen exponential growth of mobile data traffic will not be uniform across geographical areas, but is mainly concentrated in hot spots that are usually located in the built environments (BEs) such as central business districts, stations, airports, stadiums, dense urban environments, etc. This poses considerable challenges that we believe can be solved by ultra dense deployment of millimetre-wave (mmW) small-cells (SCs) in conjunction with massive multiple-input multipleoutput (MIMO) in 5G and beyond 5G (B5G) wireless networks. However, there are a number of research challenges that need to be addressed for a successful deployment of 5G/B5G wireless networks: even if the theoretical background of massive MIMO is by now rather complete, the actual performance characterization and measurements of mmW antenna arrays has not yet been fully addressed at either the component or system level; mmW radio channel measurements have been performed but with limited time delay resolution, single antennas and over single radio links; and mmW bands have been considered for mobile communications, but the level of detail and diversity of BEs necessary for meaningful mmW SC deployment has not been fully exploited. Therefore, we propose here a research approach that combines the three disruptive key enabling technologies for 5G/B5G with the aim to answer fundamental questions that are still not well understood. Hence, the research objectives of the project are as follows: • Develop and test mmW MIMO and massive MIMO antennas. • Characterize and model radio propagation channel at mmW bands for typical BEs (offices, homes, stations, airports). • Theoretically analyse and optimise massive MIMO mmW SC performance in the BEs. • Integrate massive MIMO mmW SC networks with their operating environments. • Develop methods to retrofit existing buildings and to design new buildings for efficient high-capacity wireless communications in the BEs.

Whilst traffic demand is increasing exponentially, network operators’ revenue remains flat. There is an urgent for data driven 4G/5G networks. In this project, we exploit heterogeneous big data analytics to optimize both the deployment and operations of wireless networks. We design protocols that enable future Data Aware Wireless Networks (DAWN) for enabling a new age of Internet of Everything (IoE). The proposal has been developed to address the following open issues in data driven flexible systems: • How to characterize user mobility and wireless data traffic patterns • How to infer user Quality-of-Experience (QoE) from combining data sets • How to use data analytics to assist cell planning • How to use data driven techniques to optimise the network using Self-Organising-Network (SON) algorithms • How to optimally cache data to accelerate and optimise data storage and transmission. The research objectives of the DAWN4IoE project are as follows: • Develop appropriate spatial-temporal structured filters to combine different data sets and infer both human location/mobility and digital data demand patterns. • Develop appropriate machine-learning techniques for unstructured natural language processing (NLP) to understand consumer experience for different service categories. • Design algorithms to integrate the new data analytics techniques with current state-of-the-art deployment techniques to assist HetNet planning, performance prediction, and deployment • Design mechanisms to integrate structured and unstructured data analytics to drive SON algorithms for radio resource management and smart antenna elements. • Design algorithms to optimally cache data leveraging on mobile edge computing (MEC). Achieving the above objectives will provide crucial inputs for 5G/B5G data-driven flexible wireless network design and both increase network capacity by 50% and decrease operation costs by 20-30% (compared with non-data driven networks).