Moving Networks, a Better Way to Serve Vehicular Users

Doctoral thesis, 2016

Nowadays, a great number of mobile broadband users are vehicular. It is very common that people are using their mobile devices either for working or entertainment when they are on the go. We are expecting mobile broadband to offer comparable experience everywhere, e.g., at home, in the office or on the go. One of the biggest challenges to serve vehicular users is that their communication is affected by vehicular penetration loss (VPL), which can be very high in well-isolated public transportation vehicles. This thesis investigates how to serve the vehicular users in a cost efficient way by deploying moving relay nodes (MRNs), or moving networks (MNs) on public transportation vehicles. The benefits of using MRNs or MNs are not only that they can circumvent the VPL by proper antenna placement, but also more sophisticated multi-antenna and signal processing schemes can be employed, as public transportation vehicles are less constrained by size and processing power compared to regular mobile devices. First we consider a single cell setup with one vehicular user served by the network. In [Paper A], by analytic analysis, we show that, in order to minimize the end-to-end outage probability, as the VPL increases, a half-duplex relay node needs to be deployed as close as possible to a vehicular user. This also motivates the use of an MRN to serve vehicular users. In [Paper B], we extend the study of using half-duplex MRN in several aspects, where co-channel interference, practical propagation conditions as well as inter-node handover are taken into account. From the study in [Paper B], we show that the use of MRN has great potential to improve the Quality-of-Service (QoS) for the vehicular users that are affected by moderate to high VPL. In [Paper C], we give an overview of existing solutions, and we discuss the benefits of using MRN to serve vehicular users from a system perspective. In the near future, it is expected that more frequency bands at higher frequencies will be freed up for mobile communications, especially for small cells. Therefore, in the second part of the thesis, we extend the study of using half-duplex MRNs to a full-duplex moving network (MN) in an ultra-dense urban deployment scenario. Both MRNs and MNs use wireless backhaul links to communicate with the network and form their own cells to serve the vehicular users. However, MRNs use the same frequency at their access links as the backhaul links, while MNs need dedicated frequency for their access links in order to work in a full-duplex fashion. In [Paper D], we show that the most limiting factor for further improving the performance of MNs in the ultra-dense urban scenario is the complicated inter-cell interference. Therefore, we compare the use of various multiple-antenna techniques at the backhaul receivers of MNs to alleviate the inter-cell interference experienced by the backhaul links of MNs. For the access links of MNs, as they are operating in the same frequency bands as small cells which are densely deployed along the road, we propose to use almost blank subframes (ABSs) to protect the access links of the MNs. By using system level evaluations, we demonstrate that by deploying full-duplex MNs on public transportation vehicles in an ultra-dense urban scenario, the throughput of the vehicular users can be significantly improved, and the impact on regular outdoor users is very limited. Finally, in [Paper E], we propose a novel way to enhance the uplink quality of vehicular users through cooperative communication enabled by device-to-device (D2D) communication. We show that when the vehicular users are affected by moderate to high VPL, by cooperating with each other to enhance their UL communication, the same amount of data can be sent in a shorter time. Therefore, all participants can benefit from the cooperation.