Designs Of Mathematical Model For The Energy-saving Schemes In C-RAN Based On Front-haul PONs

In recent years,with the rapid growth of smart devices,the data traffic has been erupted.The sustainability is facing severe challenges.Thus the innovation of network architecture becomes an inevitable choice.Therefore the C-RAN has been proposed,which combines the advantages of centralized BBU pool and the front-optical transportation networks(front-OTN).This thesis focuses on “The design of mathematical model for energy saving schemes on C-RAN with PON front-haul networks”,aiming to improve the resource utilization of BBU pool and optical network unit(ONU)of PON,improving the energy saving efficiency and relieving pressure of operators.These researches are valuable for the development of environmental-friendly C-RAN in the future.This thesis mainly simulates and analyzes the BBU pool and ONU of the C-RAN with PON front-haul networks from the perspectives of energy efficiency and data processing delay.And the normalized cost function is introduced to trade-off above two indexes.The main points and contributions of this thesis are listed as follows:(1)Designing of new C-RAN architecture with PON front-haul networks.The C-RAN architecture was designed by considering and integrating the logical generality of C-RAN and PON.(2)Designing the mathematical model for centralized BBU pool with dynamic energy-saving characteristics.The mathematical model is designed for the BBU pool based on the Markov chain to enable it to schedule the resources dynamically according to the tidal effect of the network data,making the least active BBUs on the premise of satisfying the specific requirement.The optimal parameter configurations are determined through numerical simulation and analysis.(3)Designing the mathematical model for ONU of front-PON networks.The Markov chain and state probability methods are used to model the operations of ONU to makes it dynamically adjusting its state according to the tidal effect of the network data.The optimal parameter configurations are determined according to specific performance requirements through numerical simulation and analysis.Finally,the comparisons about above two mathematical models are shown,and found Markov chain model is more efficient than the state probability method.This paper has completed the given design goals and achieved better results.This thesis proposed some energy saving strategies for the C-RAN network architecture,which gives important insights and practical application value for the next generation network.