Performance Optimization of Ultra-Dense Network Based on Non-Orthogonal Multiple Access

With the rapid development of science and technology, the demand for communication services has also grown rapidly. The current communication technology has developed to the Fifth Generation Mobile Communication System (5G). Ultra-Dense Network (UDN) uses a large number of low-power wireless Access Points (APs), has become a typical application scenario for 5G to improve spectrum efficiency. Non-Orthogonal Multiple Access (NOMA) can reduce the use of resource blocks and efficiently transmit signals to improve system throughput. It has become one of the most promising transmission technologies among 5G candidate transmission methods. The combination of UDN and NOMA, two technologies that are of great significance to 5G, has been the focus of scholars' research in recent years, and it is also the focus of this paper. One of the key problems of NOMA-based UDN system is how to effectively cluster APs and allocate power to serve a given user at the same frequency. To this end the specific architecture of the system has been given in this paper, considering the specific environment of the network and the needs of users, reasonably cluster the APs based on the matching theory to form an APG to assign the appropriate AP to multiple users. In contrast to prior work, the simulated annealing algorithm is exploited to re-cluster the APG on the existing clustering results and allocate the power to improve the system throughput to the maximum extent. According to a certain probability, the new scheme can accept a solution which is worse than the current solution, to jump out of the local optimal solution to get the global optimal solution. Our numerical results show that the proposed scheme has better system throughput performance than the existing schemes.