Increasing the capacity of 5G networks using mobile-cells : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Engineering at Massey University, New Zealand

Permission has been obtained for the re-use of material in journals published by The Institution of Engineering and Technology.Recently, the exponential growth in mobile data demand, fuelled by novel use-cases, such as high- definition video streaming, etc., has caused massive strain on cellular networks. As a solution, the fifth generation (5G) of cellular technology has been introduced to improve network performance through various innovative features, such as millimeter-wave spectrum, device-centric communication, and heterogeneous networks (HetNet). The HetNets will comprise of several small-cells underlaid within macro-cell to serve densely populated regions, like stadiums, malls, etc. On the other hand, due to the constant rise in the use of mobile phones while traveling, the concept of mobile-cells has emerged. Mobile-cells may well be defined as public transport vehicles (e.g., buses or trains etc.) equipped with in- vehicle cellular antenna to serve commuters. The argument for using mobile-cell is based on the observation that commuters often experience poor quality of service (QoS) due to vehicular penetration loss (VPL). Mobile-cell will decouple commuters from the core network, thus eliminating VPL, along with relieving base station off large number of users. Mobile-cells will contain multiple wireless links. Commuters will be served over access link (AL), while the communication with the core network will occur over the backhaul link (BL). On the other hand, neighboring mobile-cells will mutually exchange data over sidehaul links (SLs). Like any other device-centric communication, mobile-cells need to ‘discover’ their neighbors before establishing SLs. Neighborhood discovery is challenging for mobile-cells. Relevant literature on this topic has only focused on static devices, and discovery for mobile devices has not been investigated in detail. Hence, as our first research problem in this thesis, we have focused on the autonomous discovery by a mobile-cell. In general, due to randomness involved in an autonomous process, neighborhood discovery often fails due to collision and half-duplexing effects. This thesis focuses on mitigating these effects. Firstly, we have proposed a modified time-frequency frame structure to subside the collision and half-duplexing effects. Later on, we have presented a more reliable solution that utilizes proximity awareness to adapt transmission probability of individual devices. This scheme has resulted in a drastic increase in the probability of successful discovery as compared to the conventional approaches. On the other hand, actual data exchange via mobile-cell’s links requires interference-free resource allocation for each link. Mobile-cells’ wireless links will cause severe interference to the out-of-vehicle cellular users. Few researchers have assigned separate bands for in-vehicle and out-of-vehicle links. However, given the scarcity of spectral resources, these methods are practically inefficient. Thus, we have addressed the issue of resource allocation as the second research problem in this thesis. Instead of assigning individual resources to each link, we have focused on resource sharing between multiple wireless links. To achieve this goal, we have exploited VPL and utilized successive interference cancellation. Our results have shown high QoS at each individual link. We have also demonstrated the effect of mobility on the proposed resource sharing schemes. The schemes proposed in this thesis will ensure that the mobile-cell increases the capacity of 5G networks through aggressive resource sharing such that more links will use available spectral resources