Network services in the future will be highly heterogeneous and differ substantially in their requirements on the underlying communications network. There will be a need for fast and reliable connectivity with ultra-low latency for use cases such as remote control of robots, and support for billions of sensors and things. 5G will also need to provide consistent and high quality connectivity for people and things.
To meet the requirements of such heterogeneous services, networks with different topologies will be needed e.g. to achieve low latency for automotive applications network functions should be placed at the network edge, while a central placement of network functions allows to exploit pooling gains with mobile broadband services.
However, building separate networks per service is highly inefficient from an economic viewpoint. To overcome this problem, 5GPP aims to design a novel 5G mobile network architecture which provides a higher degree of flexibility in order to accommodate use cases which are going to be relevant for 5G mobile network deployments based on network slicing. Network slicing allows for operating multiple logical networks with their own topologies on a common infrastructure.
In this project we are planning to build an end-to-end 5G flexible (test and development) platform while mainly focusing on novel architectural aspects of 5G networks and also some of the enabling technologies upon which it relies e.g. Software-Defined Networking (SDN) and Network Function Virtualization (NFV), Multi-Connectivity and Flexible RAN within the first year of implementation, whereas during the second year we will be dealing with the development of 5G-specific physical layer features (e.g. mmW and massive MIMO) and their integration into the platform.