The importance of Heterogeneous Networks (HetNets) is increased after enabling the high band spectrum such as millimeter Wave (mmWave), in the development of the Fifth Generation (5G) wireless communication system. In a HetNet, small cells are suitable to operate by using mmWave due to its short-range, while macrocells are liable to use long-range radio waves. As a result, the performance of the next generations of wireless communication systems will enhance dramatically. However, the networks' architecture has become more complex and challenging to manage and optimize. Besides, the handover (HO) among small cells is another big challenge and needs to address it on a prior basis. Different techniques are proposed in the literature to improve the current network architectures. In this paper, we provide fundamental background concepts and notions used in the 5G wireless communication system. Besides, we study different HO techniques and management schemes that perform acceptably in the dynamic nature of next-generation wireless networks. Finally, the software-defined network and machine learning-based approaches are suggested as solutions for HO management in the 5G HetNet system.