Research On User Association And Resource Allocation Mechanism In TWDM-PON And C-RAN Joint Architecture

Cloud radio access network(C-RAN)uses time and wavelength division multiplexed passive optical network(TWDM-PON)as the fronthaul,effectively combining optical and wireless networks It has the advantages of providing users with high-reliability and large-capacity network access services,while reducing deployment costs and operating costs,and providing flexible network expansion services.However,the explosive growth of various new services has brought huge challenges to network performance in mobile wireless networks.Therefore,effectively utilize the network resources in the TWDMPON and C-RAN joint architecture,meet the traffic needs of users,and maximizing the efficiency of network resources is the focus of future research.This thesis fully takes into account the advantages of the joint architecture of TWDM-PON and C-RAN,guarantees the quality of service(Qo S)to meet the user experience,makes full use of network resources,and further optimizes the network throughput.A load balancing user association mechanism and a service priority aware resource allocation mechanism are proposed.Firstly,this thesis introduces the research background,access network architecture and development trend of optical and wireless combined network based on TWDM-PON and C-RAN.Among them,the research hotspots in the optical and wireless combined network based on TWDM-PON and C-RAN are emphatically introduced,and various network resource allocation mechanisms in the TWDM-PON and C-RAN joint architecture are summarized.Secondly,the distributed remote radio head(RRH)load imbalance in the wireless network limits the transmission efficiency of the TWDM-PON and C-RAN joint architecture.This thesis proposes a load-aware user association mechanism to make full use of the network resources of the TWDM-PON and C-RAN joint architecture and ensure user Qo S.A revenue function is constructed according to the rate requirements of different users and the impact of RRH load on users.On the basis of ensuring the user request rate,a random game model is established according to the load status of the RRH in the network,and a load-balanced user association and resource allocation algorithm is proposed based on multi-agent Q learning,so as to obtain relatively optimal user association and resource allocation strategy.After simulation analysis,the proposed load balancing user association mechanism can achieve load balancing between RRHs in the wireless network on the basis of guaranteeing user Qo S,and improve network throughput.Thirdly,effective resource management strategies can efficiently use network resources,ensure the reliability of network services,and improve the utility of resources in the network.This thesis proposes a Service priority aware resource allocation mechanism in the TWDM-PON and C-RAN joint architecture.In response to the problem of mismatch between resource allocation and service Qo S under high load conditions,the weights of optical network unit(ONU)and wavelengths are defined according to the Qo S requirements of different services,and the ONU wavelengths are adjusted for resource allocation according to the weight model.Then determine the utility of the base band unite(BBU)pool based on the computing resources obtained from the wavelength,and turn resource allocation into a problem of maximizing the total utility of computing resources in the BBU pool.A Service priority aware resource allocation algorithm based on Q learning is proposed to get a relatively optimal wavelength allocation plan.