Research On 5G Cloud Access Networking Technology And Resource Management Methods

With the upgrading of mobile communication system,high-speed connected 5G era is coming.Compared with the existing 4G,the 5G network pays more attention to the unmet 4G challenges.Higher capacity,faster data rate,lower end-to-end delay,lower overhead,large-scale equipment connection and consistent user experience quality are the goals of 5G communication system.At present,a large number of base stations lead to high energy consumption,the CAPEX/OPEX of network is increasing year by year,and the "tidal effect" results in low base station utilization.In this scenario,the access network faces the challenge of high capacity and low latency with low-cost transmission.In order to meet the demand of high capacity and low power consumption of access network under the condition of "tidal effect",this paper adopts fronthaul bearer scheme based on TWDM-PON and adjusts the bandwidth of fronthaul network with user traffic awareness.In addition,we also do BBU pooling and baseband resource management research work.The main research work is as follows.Firstly,comb and contrast the current typical fronthaul network schemes in depth,including the advantages and disadvantages of each scheme in terms of bandwidth efficiency,transmission distance,delay,reliability,networking,cost,scalability,and the respective applicable scenarios.The fronthaul network solution based on TWDM-PON can provide flexibility for users to adjust bandwidth,share network resources and expand services under the premise of ensuring high bandwidth.It has the advantages of low cost in scenarios of existing PON networks and meets the 5G fronthaul requirements of large capacity and low consumption.It is an ideal 5G fronthaul program.Secondly,aim at the problem of bandwidth linited brought by the large capacity demand in 5G mobile cloud access network,a traffic aware fronthaul bandwidth on-demand adjustment algorithm and implementation plan is proposed.Based on TWDM-PON fronthaul network,real-time monitoring obtains the load of each RRU(Radio Remote Unit),and periodically adjusts the RRU wavelength and bandwidth template by traffic awareness algorithm to dynamically adapt to traffic changes to achieve high network resource efficiency purpose.The simulation results show that in the "tidal effect" scenario,compared with the traditional fixed wavelength allocation scheme,the bandwidth utilization rate can be increased up to 75%and the network throughput can be increased up to 44%.Thirdly,in the BBU resource pooling scenario,a dynamic optimal fit Decreasing(DBFD)algorithm that dynamically allocates BBU resources according to the RRU traffic is proposed.RRU signals are rationally allocated under the premise of meeting RRU transmission requirements and not exceeding BBU processing capabilities.The scheme makes full use of the BBU resources and minimizes the number of enabled BBUs to solve the problem of high energy consumption caused by idle running of a large number of BBU resources in a dynamic traffic scenario.The simulation results show that the proposed scheme saves more than 70%of the energy consumption under the scenario mentioned in the thesis,compared with the traditional allocation scheme of fixed RRU-BBU mapping relationship.Forthly,in order to verify the feasibility of bandwidth adjustment of 5G fronthaul based on SDN(Software Defined Network)network architecture,I participated in the construction of a software-defined optical access network experimental platform based on OpenDaylight.Because there is no commercially available TWDM-PON with open interfaces,we use controller,Openflow switches and EPON devices in this project to experiment on the rate limiting of traffic and to simulate the scenario of dynamically adjusting resource allocation based on network traffic in the fronthaul network.Simulation results show that based on SDN network,we can achieve reasonable bandwidth allocation and limit traffic rate for different users according to network needs.