Traffic-Aware Network Slicing for 5G Networks in Cloud Fog-RAN over WDM Architecture

Fifth generation (5G) mobile networks will utilize network slicing to support a diverse range of services, each with distinct specifications for data rate, latency, and reliability. These services are broadly classified into 3 main categories: ultra-reliable low-latency communication (URLLC), enhanced mobile broadband (eMBB), and massive machine type communication (mMTC). Network slicing efficiently provisions these services by creating multiple virtual networks over the same physical infrastructure using network function virtualization (NFV). In this paper, we propose a traffic-aware network slicing scheme by levaraging a 3-layer cloud fog radio access network (RAN) (CF-RAN) over wavelength division multiplexing (WDM) architecture. This network architecture consists of remote radio heads (RRHs) at the first layer, fog nodes at the second layer, and baseband unit (BBU) hotels at the third layer. The fronthaul for incoming URLLC requests are served at the fog nodes because of their stringent latency requirements, while the fronthaul for eMBB and mMTC requests is handled at BBU hotels. The proposed architecture dynamically adapts to the random traffic generated by each RRH. Furthermore, we propose a heuristic algorithm to investigate how changes in the service request ratio effects the overall performance of the proposed architecture. The algorithm minimizes resource utilization while maintaining maximum network connectivity. The results demonstrate the impact of incoming slice request variation on the number of active nodes, number of served requests, and aggregated fronthaul delay.