Heterogeneous architecture is an underlining feature of 5G, however deployment and management of HetNets in 5G scenarios is yet to be explored. Given the need to satisfy overwhelming capacity demands in 5G, mm-wave spectrum (3-300 GHz) is expected to offer a very compelling long term solution by providing additional spectrum to 5G networks. Hence, the challenge is the integration of mm-wave in heterogeneous and dense networks as well as the backward compatibility and integration with legacy 4G/3G networks. Furthermore, Cloud radio access networks (C-RAN) contribution to 5G is considered as a cost effective and energy efficient solution for dense 5G deployment. From an energy point of view, cost and energy consumption are major considerations for 5G. C-RAN and energy efficiency techniques could help in performance improvements.

Although HetNets were introduced in 4G networks, their complexity has increased in 5G networks. In this paper, we will try to build a clear image of HetNets in 5G cellular networks. We consider different technologies with a special focus on mm-wave networks given its important role in 5G networks. We then address the available standards in HetNets that allow interworking and multihoming between different radio access technologies. Afterwards, we consider the virtualization of 5G HetNets and its benefits. Different resource allocation strategies in the literature are also presented for single-resource as well as for multi-resources. Finally, we give an overview of existing works addressing energy efficiency strategies in 5G networks.

*Content Disclaimer Note (Added by Committee): This communication is a preprint uploaded under author responsibility. The Congress committee, only do a preliminary inspection of topic suitability. The content of this preprint communication is responsibility of authors and do not express the opinion of the members of committee. The committee is not responsible from content veracity or originality. Using automatic text generation tools, like ChatGPT, is allowed only for AI software/script coding purposes or as a way to improve quality of redaction of the original text. We recommend using text similarity analysis tools but this an author's decision.