Deployment Optimisation of a 5g Heterogeneous Network.

Abstract

Non-orthogonal multiple access (NOMA) has been considered one of the key enabling technologies of 5G networks. This has been mainly due to its high spectral efficiency capabilities since it enables multiple users to simultaneously utilise the same spectral resources. On the other hand, the evolution towards 5G has been supported by the high data demands and the need to efficiently utilise spectral resources. Thereby, NOMA has been one of the suggested solutions. In this research, a two-tier NOMA-based HetNet has been analysed to optimally place a set of small BSs. Simulated annealing (SA), a meta-heuristic optimization algorithm, has been utilized to optimize the network by proposing approximate optimal candidate locations. Thereafter, a placement algorithm has been proposed to select the optimal set of locations viable for the deployment of small cells. This proposed approach is compared with a random deployment approach in which an equal number of small BSs are deployed randomly from a set of optimised candidate sites. The aim of optimisation in the research is to select optimal locations of the small cells as well as minimise the number of deployed small cells. Therefore, the minimization objective calls for strategic deployment of the small cells. Given their small size, there are many possible deployment locations. Therefore, the objective of this research is to find the optimal locations of the small cells in a dense 5G heterogeneous network. In essence, if small cells are placed in optimal locations, it implies that the number of required BSs to reach particular performance objectives will be minimized. Hence, the ability to manage the soaring interference and the energy consumption emanating from the network densification. To test the efficiency of the used approach, a series of simulations have been conducted using MATLAB. The simulation results show that the proposed approach significantly improves the sum rate and coverage of users by 22% while managing to keep the interference below the set threshold. The optimal deployment strategy has been proven to significantly enhance performance. For instance, the proposed approach has proved effective in the selection of nearly optimal locations for the deployment of small cells. Furthermore, the nearly optimal small base station (BS) densities have been determined. Beyond the optimal density, further small BS deployments will have no significant effect on improving the network parameters. In addition, the proposed algorithm has proved energy efficient with about 0.23 bps/Hz/J extra hence providing a high system capacity