Research On Fault Tolerance Mechanisms In Fiber-wireless Access Network

As effectively realizing the combination of high capacity,long distance in the optical access network and the low cost,high flexibility in the wireless access network,the FiberWireless(FiWi)access network is considered as one of the main solutions of the next generation broadband access networks.It can be seen that FiWi can provide strong access and transmission capabilities for a large number of users.However,due to natural disasters or human factors,once a failure occurs in FiWi access network,it will cause a large number of traffic interruptions.Then the availability of traffic is seriously affected and the quality of user experience is decreased.Therefore,the fault tolerance mechanisms in FiWi access network has important research significance.The fault tolerance mechanisms in FiWi access network are studied in this dissertation.According to the characteristic of FiWi access network,that is various types of failures,complex topology,variable network environment,heterogeneous transmission media,etc.,the objective of this dissertation is improving network survivability and ensuring service availability in the FiWi access network.This dissertation has carried out in-depth research on dynamic path switching,fast data recovery,and efficient resources deployment to enhance the capacity of tolerating intra-segment failure and segment level failure.This dissertation is trying to build a next-generation broadband access network with strong capability of fault tolerance.(1)Dynamic path switchingTo slove the issue of traffic transmission quality deterioration which is caused by network congestion,the internal factors that deteriorate the transmission quality during traffic recovery are analyzed and a mechanism called Efficient Traffic Recovery Mechanism with Load Balancing(DTRM-LB)is proposed.Meanwhile,a traffic grooming model with a large number of traffic recovery was established.Considering the current state of network resource from global perspective,a bipartite graph model with node segmentation is proposed to achieve reasonable gateway resource configuration.Meanwhile,aiming at the paths are prone to aggregate when recovery for a large number of traffic,a path allocation scheme with congestion avoidance is proposed to dynamically allocate high transmission recovery path.Numerical results indicate that the DTRM-LB mechanism can effectively balance the network load.Meanwhile,the problems of delay and packet loss rate are decreased which are caused by network congestion.Then the traffic transmission quality is guaranteed.(2)Fast data recoveryTo slove the issue of massive packets loss in recovery,the cross-domain collaborative and reliable transmission mechanism is studied.Furthermore,a mechanism called Lossless Fault-Tolerance Mechanism Combining with Parallel Routing and Network Coding(LFTM-PR-NC)is proposed.By utilizing the information fusion of network coding and high reliability of parallel paths,a cross-domain collaborative parallel transmission model is established and the proposed mechanism breaking through the normal thinking of traffic recovery.LFTM-PR-NC mechanism can tolerate single path failure which is caused by network failure by integrating traffic recovery with route coding.Numerical results indicate that the LFTM-PR-NC mechanism can greatly reduce the recovery time of interruption traffic and improve the transmission delay.(3)Efficient resource deploymentTo slove the issue of highly backup resource deployment cost and long recovery time,a mechanism called Backup Fiber Deployment Algorithm Based on Traffic Complementarity(BFDA-TC)is proposed and a backup resource deployment model at time and space scale is established.In order to improve the recovery speed,a remote node with backup protection switching is designed to eliminate bottleneck links in the recovery process.In addition,focusing on the complementary characteristics in traffic demand caused by the social life of human,the dynamic change rule of idle resources at the time scale and dynamic game rule between the cost of backup resource deployment and recovery speed at spatial scale is studied.And then the backup resource is deployed between the segments which have significant complementarity characteristics of traffic demand by combining minimum cost maximum flow theory in the flow network.In addition,the idle resource is sharing more granularly.Furthermore,the cost of backup resource and the recovery speed are decreased.Numerical results indicate that the BFDATC mechanism can decrease the cost of backup resource in most scenarios and the recovery time is effectively reduced in any scenarios.