Difference Based Matching Algorithm for User Association Problem in Ultra-Dense Heterogeneous Networks

To meet the ever-increasing requirements of high data transmission rates by Internet of Things (IoT) users, ultradense heterogeneous network (HetNet) is introduced for its excellent performance in enhancing cell coverage and improving system throughput. However, the sophisticated structure in ultradense HetNet makes user association become a crucial problem. Thus, this paper focuses on the user association problem in ultradense HetNet, aiming at maximizing system throughput. Then a difference based matching algorithm is proposed, in which the achievable rate difference between the current request and the next request is taken into account when multiple IoT users compete for the access opportunity with the same low power points (LPPs). Simulation results show that the proposed algorithm can achieve both the highest system sum-throughput and the best fairness compared with the four comparison algorithms.