Research On Collaborative Optimization For Optical And Wireless Resources In Radio Access Networks

With the development of Fifth-generation and beyond mobile communication(5G/B5G),mobile networks are dedicated to accommodating more subscribers and diversified service scenarios.5G/B5G is expected to significantly enhance the quality of experience on services,and pave way for the interconnection of everything.The radio access network(RAN),as the front end of mobile networks,plays an important role in the 5G/B5G evolution.The optical network is then introduced for data transmission between the base station and core network because of its high-capacity and low-latency features.Thus,RAN will evolve towards a converged optical-wireless network,where multiple resource dimensions are involved.However,the current RAN is facing the issues of low resource efficiency as well as high CAPEX/OPEX(Capital Expenditure/Operating Expense)and power consumption.Essentially,the above issues arise from the low adaptation efficiency of wireless and optical resources.Thus,how to improve the adaptation efficiency of heterogeneous resources is a challenging problem in RAN.To address the above issues,this paper focuses on the resource dimensions of "wireless transmission","baseband processing",and"optical transmission",and then solves two problems of "how to achieve the high-efficiency adaptation between wireless and optical transmission resources",and "how to collaborate the baseband processing and optical wavelength configuration." Through the studies in this paper,we hope to improve the resource efficiency of RAN and thus reduce the network expenditure and power consumption.The main contributions and innovations of this research are concluded as:(1)Research on elastic fronthaul networking and integrated resource optimization for multicast-service beam-forming in massive MIMO enabled RAN.The integration of massive MIMIO beamforming and optical networks is facing the problems of heterogeneous resources adaptation and transmission resilience.To address these issues,this paper proposes a flexible and resilient fronthaul network architecture,in which any antenna can be mapped or switched to any optical wavelength,thus remarkably enhancing the network resilience.Then,an optimization strategy for multicast-service beamforming is proposed,where antennas,resource blocks,and wavelengths are jointly allocated.This strategy is dedicated to incorporating the same services into the identical beam groups so that significantly reduces the fronthaul bandwidth requirement.The simulation results show that low dropping probability can be achieved with our proposed fronthaul architecture even if several wavelengths are in failure.Moreover,for a large number of subscribers in networks,our proposed strategy can contribute to the fronthaul bandwidth(≥50%)saving at the expense of few extra antennas or resource blocks(≤10%).(2)Research on baseband processing deployment and optical transport mechanism with the fine-grained functional split.There is a contradiction between the centralization gain of baseband processing and transport resource-saving in radio access networks.To address this issue,a flexible baseband processing deployment and optical transport mechanism with the fine-grained functional split is investigated.The traditional baseband unit(BBU)is divided into several fine-grained units(FU).Then,a quantitative model for estimating computational complexity and bandwidth requirement is established.This paper also proposes an optimization strategy for a resource-efficient FU deployment,which aims to minimize the number of activated processing pools,bandwidth consumption,transport latency,and network cost.Simulation results show that fine-grained split can benefit the centralization gain,bandwidth(30%～50%),and cost saving(5%～20%)compared with traditional C-RAN and recently-emerged NG-RAN architecture.But fine-grained split may increase the transport latency because of distributed function processing and frequent data switching.Our analyses provide insights into the modeling and design of efficient radio access networks in 5G and beyond.(3)Research on energy-efficient DU-CU deployment and lightpath provisioning strategy.To solve the power saving problem in diversified services enabled NG-RAN,an energy-efficient DU-CU deployment and lightpath provisioning strategy is proposed.In this strategy,a power model for both DU-CU processing and optical transmission devices is established.The network devices can be turned from "always-on" to "on-demand on" pattern to support the power saving.Then,the DU-CU locations and lightpath configuration are decided through this strategy,which minimizes the power consumption of networks.Numerical results show that our proposed strategy benefits the power reduction under different service scenarios.Moreover,our strategies can also be applied with adjustable deployment that adapts to time-varying traffic,which reduces the power compared with conventional static deployment(reduce 10%power under 120 AAUs scenario).Our analyses can help provide insights into the modeling and designing of energy-efficient 5G/B5G networks.