Resource Allocation Optimization for Time and Wavelength Division Multiplexing Passive Optical Network Enabled Mobile Fronthaul With Bitrate-Variable Compressed Common Public Radio Interface

Abstract

Recently, a cloud radio access network (C-RAN) has been proposed as a candidate architecture for fifth-generation mobile communication. Fronthaul is a new segment in C-RAN. In this paper, we investigate the algorithms for resource allocation in time and wavelength division multiplexing passive optical network enabled front-haul. We formulate an integer nonlinear programming (INLP) model, considering the operation mode of the small cell. A heuristic method is also proposed based on an adaptive parallel genetic algorithm (GA). Three optimization objectives are considered when we implement the resource allocation schemes in the fronthaul: 1) minimize the total number of used wavelengths, 2) minimize the load imbalance of the fronthaul, and 3) minimize the amount of traffic influenced by fronthaul virtual topology change. The results show that INLP and adaptive parallel GA have good performance for resource allocation in the fronthaul. Wavelength resources can be significantly saved in different load scenarios, so that the load difference can be minimized with minimal topology adjustment. Furthermore, the adaptive parallel GA obtains more efficient resource allocation than the traditional GA with much less running time, which makes it applicable for the large-scale fronthaul network optimization with fast convergence.