基于层次回声状态网络的电力EPON业务流感知

为了进一步提高电力EPON（Ethernet Passive Optical Network）对智能电网多业务的支持能力，文章提出了一种层次基于回声状态网络的电力EPON业务流感知技术，采用离散回声状态网络算法对电力通信业务流进行识别感知，以实现EPON对业务的高效匹配能力。在此基础上通过OLT与ONU的交互，优化电力EPON的资源调度与分配能力。为了验证所提出的业务流感知机制，以配用电通信EPON作为典型业务，对基于回声状态网络算法的业务感知进行仿真，仿真结果表明该机制能有效降低丢包率和缩短时延，提高了EPON系统对电力通信多业务的支持能力。

Hierarchical Echo-State-Network Based Sercice-Traffics Awareness in EPON for Smart Power Grid

Aimed to improve the ability of EPON (Ethernet Passive Optical Network) to support multiple services brought by smart power grid, a Hierarchical Echo-State-Network Based Traffic-Awareness Mechanism in EPON system is proposed for energy internet. The proposed mechanism is able to be aware of characteristics of multi-services and to match their requirements with better quality, through the echo-state-network algorithm. Thus, resources allocation optimization of EPON system is enabled by the proposed mechanism through cooperation between OLT and ONUs. Simulation evaluation of this proposed scheme is conducted and takes smart power grid services as typical application. Simulation results show that this proposed approach can greatly improve the supporting ability for multiple services of smart power grid, in terms of packet loss rate and time delay performances.