The Performance of Dense and Heterogeneous LTE Network Deployments within an Urban Environment

Traffic in the mobile broadband networks is expected to grow very rapidly in the coming years. This traffic growth is caused both by the evolution of mobile terminals and by the increased use of more traffic-heavy services, such as video. In order to be able to meet the increased capacity needs, the existing mobile networks have to be densified, either by deploying new macro sites, or by deploying new low-power sites within traffic hotspots. This doctoral dissertation provides an overview of a few different network densification alternatives and compares their performance and energy-efficiency with the help of advanced radio network simulations. In addition, the impact of different network design choices is evaluated. The results demonstrate that the heterogeneous network deployments are realistic alternatives to the traditional way of densifying mobile networks by deploying new macro sites. However, the price to pay is that a considerably larger number of new sites will be required to obtain the same network performance. Heterogeneous network deployments can be made more efficient by increasing the output power of the low-power eNodeBs or by carefully planning the locations of the low-power sites so that the obtained level of the traffic offloading can be maximized. The traffic offloading can be increased also with the help of biased cell selection, but in that case the quality of the downlink control signaling can become the limiting factor unless some form of enhanced inter-cell interference coordination mechanisms are applied at the same time. The obtained results indicate that the densified macro deployment is in many cases the most energy-efficient network densification alternative. However, if some form of fast cell DTX is applied to idle cells, heterogeneous network deployments become much more competitive since the cost of fixed power consumption can be reduced. The energy-efficiency of densified network deployments can be enhanced also by switching off underutilized capacity cells, or by switching idle capacity cells to sleep mode. Finally, design choices aiming to reduce the required number of low-power cells are shown to be beneficial also from the network energy-efficiency point of view.