Research On Delay Model And Algorithm Of Traffic Grooming In Deterministic Optoelectronic Hybrid Networks

Ultra low latency communication is an important scene of the 5th generation(5G)mobile communication,which puts a serious challenge to the 5G bearer network.Ethernet-over-WDM is one of the networking schemes of 5G bearer networks,with low cost and high bandwidth.The latency-sensitive service requires the optoelectronic hybrid network to have deterministic latency transmission performance.Traffic grooming is important technology that improve bandwidth resource efficiency,however,traffic grooming leads to random queuing delays in packets,urgently need to establish a latency-bound model for traffic grooming to ensure deterministic transmission requirements for latency-sensitive services.This paper studies the latency model and algorithm of traffic grooming in deterministic optoelectronic hybrid network,and research work and innovations are as follows:(1)In view of the "grooming delay" problem in the traffic grooming process,in this paper the latency-bound model of traffic grooming is established based on stochastic network calculus.Taking the end-to-end latency upper bound of the service as the optimization goal,the latency-bound model is applied to the auxiliary graph model,and a low-latency bound dynamic traffic grooming policy based on the auxiliary graph model is proposed.Compared with traditional traffic grooming policies,the simulation results show that the proposed policy can reduce the end-to-end latency bound of latency-sensitive services,and select the path with a low-latency bound for transmission.At the same time,it indicates that the policy can further reduce the end-to-end latency of the latency-sensitive services when the frame preemption schedule mode is selected.(2)Aiming at the problem of limited frame preemption function of nodes,this paper proposes a traffic grooming auxiliary graph model based on frame preemption sensing,and designs a dynamic traffic grooming policy according to the auxiliary graph model.Simulation results show that the policy proposed in this paper can select the best path under frame preemption constraints to realize low-latency transmission of latency-sensitive services.At the same time,it shows that as the number of nodes with frame preemption function in the network increases,the delay of latency-sensitive services decreases.