A Smart QoE Aware Network Selection Solution for IoT Systems in HetNets Based 5G Scenarios

As small cells and network Access Points (APs) have been massively deployed in the cities, industrial, academic campuses, communication techniques utilising same or different spectrum bands are required to coordinate each other and coexist in harmony. Meanwhile, User Equipments (UEs) including Internet of Things (IoT) devices are more advanced in terms of communication ability. Simultaneously supporting data transmission through multiple network interfaces is prevalent nowadays. The current networks have many diversities and will continually involves more user scenarios and more advanced smart devices. The communication between those devices, is mostly through wireless media, such as, bluetooth, WiFi or LTE. Meanwhile, with the evolution of wireless technology, the communication in most systems will migrate to Heterogeneous Networks (HetNets)/5G networks. The devices in those systems will be able to support 5G communication, enabling more than one radio access interfaces through Multiple-In-Multiple-Out (MIMO) antennas. In order to realise 5G communication, massive regulations and standards are proposed and HetNets have become one popular means since it can utilise the existing infrastructure. The focus of 5G has changes since the last generation from operator centric to user centric services. For purpose of achieving Quality of user Experience (QoE) supported 5G, network selection in HetNets based approach is necessary and also should be intelligent. In this paper, we have proposed that the classic utility function can be applied in such a HetNets scenario. With modifications on the classic utility approach, more advanced features can be enhanced to support QoE aware communication in HetNets based 5G networks. Furthermore, we have demonstrated this idea through a LTE and WiFi integrated network. A solution for smart network selection is delivered in order to achieve 1) better user experience, 2) manageable resource allocation and 3) price control. Based on the analysis, we argue that such a solution can also be extended and implemented for more complex 5G networks to accomplish QoE awareness.