On the use of unmanned aerial vehicles to strengthen and assist cellular networks

Abstract

This thesis addresses the open research problem of efficiently use Unmanned Aerial Vehicles to mitigate and eventually recover Cellular Networks in extreme conditions, such as congestion or natural disasters. Small scale UAVs are at a mature stage of development and can today carry lightweight commercial micro-cells with smaller form factor. The mobile micro-cell can be used to offload users from a congested micro cell to optimize the bandwidth usage for applications such as video streaming. The delay mitigation problem for video streaming applications in congested cellular macro-cells by using a mobile micro-cell mounted on an UAV is studied. Firstly the thesis discusses algorithms and comprehensive design criteria for user offload selection (selecting what users need to be offloaded to the micro-cell) and drone positioning (selecting the position of the UAV that minimizes the network delay). The effectiveness of the proposed criteria is evaluated through extensive performance evaluation. We show that the performance increases consistently in terms of bandwidth requests mitigation and average delay reduction under different system configurations. Secondly a prototype of an Aerial Base Station compatible with 5G architecture is proposed. The proposed framework allows for on-the-fly drone repositioning based on rigorous optimization techniques using real-time network metrics to enhance users' service. Experimental results demonstrate that UABSs are able to autonomously reposition themselves based on cellular network metrics and to improve network performance. Finally the HIRO-NET robotic framework is presented. HIRO-NET is an emergency infrastructure-less network tailored to address the problem of providing connectivity in the immediate aftermath of a natural disaster, where no cellular or wide area network is operational and no Internet access is available. HIRO-NET establishes a two-tier wireless mesh network where the Lower Tier connects nearby survivors in a self-organized mesh via Bluetooth Low Energy and the Upper Tier creates long-range VHF links between autonomous robots exploring the disaster stricken area. In the event of an Internet connection still being available to some user, HIRO-NET is able to opportunistically share and provide access to low data-rate services (e.g. Twitter, Gmail) to the whole network. Results show that the Lower Tier always reaches network convergence and the Upper Tier can virtually extend HIRO-NET functionalities to the range of a small metropolitan area.--Author's abstract