Multi-dimensional resources allocation based on reconfigurable radio-wavelength selective switch in cloud radio over fiber networks.

To support the bandwidth-hungry applications and ubiquitous services, cloud radio access network (C-RAN) provides the large bandwidth connecting hundreds of terminals for different demands. Applying the technology of radio over fiber (RoF), cloud radio over fiber network (C-RoFN) migrates the processing and calculation functions of the base station to processing unit (PU) which simplifies the system complexity and reduces the data flood caused by digitization in 5G and beyond. The flexible allocation of multi-dimensional resources including radio and optical resources is necessary for C-RoFN to enhance the resource utilization and the quality of services. However, there is a lack of effective method to adjust spectral resources occupied by the RoF signal in C-RoFN. In our previous work, we realized multi-stratum resources optimization between processing and optical resources for C-RoFN from the perspective of network layer. In view of this, this work extends to consider the elastic switching architecture in the physical layer of C-RoFN and the scheduling of radio and optical spectral resources based on the proposed architecture. We propose a novel flexible switching and scheduling mechanism for multi-dimensional resources allocation in software defined C-RoFN. The main contributions on this work are twofold. 1) According to the flexible control demand of multi-dimensional resources, a reconfigurable radio-wavelength selective switch (RWSS) architecture is designed in a C-RoFN scenario to realize the tunability of spectral resources occupied by the RoF signal in optical network layer, which aims to break the restrictions of wavelength consistency and wavelength conflict. 2) Based on the proposed RWSS architecture, a routing, radio and wavelength assignment (RRWA) algorithm considering wavelength-frequency unsaturation is provided to implement the flexibility and allocation of radio and optical spectral resources. The feasibility and efficiency of the architecture are demonstrated on the testbed including date plane using software defined optical and wireless nodes experimentally. The performances of RRWA algorithm under heavy traffic load scenario are also quantitatively evaluated compared with the other algorithm in terms of resource utilization rate, blocking probability, and provisioning latency.