Title:
Capacity Dimensioning for 5G Mobile Heterogeneous Networks
Supervisors
Assoc. Prof. Henrik Christiansen
Evaluation Board
Assoc. Prof. Sarah Reneé Ruepp, DTU Fotonik
Member of Technical Staff, Small Cell Research, Anna Wielgoszewska, Bell Labs, Nokia, Ireland
Prof. Klaus Pedersen, Aalborg University, DK
Master of the Ceremony
Assoc. Prof Henrik Wessing, DTU Fotonik
Abstract:
Increasing the network capacity has historically been the main driver for the evolution of mobile communications, and that is no exception for the fifth generation (5G) of mobile networks. Therefore, enhancing the mobile broadband communications is one of the purposes of 5G.
This thesis analyses the performance and interworking of three solutions that have arose as key pieces for boosting the capacity in 5G: heterogeneous networks (HetNets), millimiter wave (mmWave), and non-orthogonal multiple access (NOMA). Through diverse network scenarios and simulation models some of the challenges that are encountered when implementing mmWave and NOMA in HetNets are addressed, and methods to approach such challenges are proposed and evaluated.
Performance analyses provided for single cell and network scenarios show that the use of mmWave and NOMA can help 5G neworks reach multi-Gbps rates and high spectral efficiency. Nevertheless, both technologies require rethinking how some of the network functions are done, and their implementation is not trivial. For mmWave, the challenges associated to the initial cell search in small cells are addressed; showing that new search methods are required to adjust to the highly directional transmissions that characterize communications at these frequencies. Approaches for HetNets where the macro and small cells cooperate to improve the performance of said methods are proven to be beneficial from both the access delay and a capacity dimensioning perspective.
For NOMA, special emphasis is done in how to mitigate the effects of the intra-cell interference. From the simulations results it is shown that such interference can diminish the capacity benefits of NOMA if adequate methods are not used to account for it, especially in scenarios where NOMA coexist with other multiple access schemes. An approach considering transmission parameters adjustments done at the base station for NOMA is proposed; showing that with such type of adjustments the network can significantly benefit from the high capacity that NOMA offers. The importance of the user pairing method in the performance of NOMA is also addressed, confirming that the increase in the network capacity is directly related to the user-pairing algorithm. Moreover, it is shown that a flexible selection of the pairing algorithm subject to the load conditions in the network is preferred, since it can provide a balance between the algorithm complexity and efficiency and the network capacity. Therefore, this is a topic that requires special planning when considering the use of NOMA, and even more so when mmWave is used.
For HetNets deployments, it is shown that the capacity can be further increased with NOMA when techniques that aim at maximizing the NOMA UEs and minimizing the OMA UEs in highly loaded cells are used; in this regard, the implementation of load balancing techniques designed for NOMA is suggested, since they can result in a better utilization of the available resources in the network.
Furthermore, through capacity dimensioning calculations it is shown HetNets with NOMA and mmWave are a significant part of the solution for coping with the growing demand for data services and help 5G network reach their capacity goals.