论文部分内容阅读
光纤网络中受到复杂的电磁色噪声干扰,云数据调度在干扰下容易产生资源蜕变,导致调度实时性和准确性不好。传统方法采用优先级列表调度方法,在变异尺度调整中产生时间漂移,影响调度准确性。提出一种基于自相关匹配滤波和时分多址时隙分配的光纤网络中云数据分簇调度算法。对光纤网络中噪声干扰进行自相关匹配滤波抑制,进行了云数据调度模型的总体设计和云数据特征信息流提取,对提取的光纤网络噪声干扰下的特征信息流进行自相关匹配滤波处理,采用时分多址时隙分配方法实现滤波后的云数据的优化调度,按照前导时隙的分配机制,使待调度的光纤网络云数据特征形成最佳匹配。仿真结果表明,该模型进行云数据调度的抗噪性能较好,云数据的特征匹配度较高,调度准确度较高。
Fiber network is subject to complex electromagnetic noise, cloud data scheduling prone to resource degradation under interference, resulting in scheduling real-time and accuracy is not good. The traditional method adopts the priority list scheduling method, which causes time drift in the variation scaling and affects the scheduling accuracy. A cloud data clustering scheduling algorithm in optical fiber networks based on self-correlation matched filtering and time division multiple access slot assignment is proposed. Autocorrelation filtering is used to suppress the noise interference in the fiber network, and the overall design of the cloud data scheduling model and the extraction of cloud data feature information are carried out. The feature information stream extracted from the optical fiber network is subjected to autocorrelation matching filtering. The time division multiple access slot allocation method realizes the optimal scheduling of the filtered cloud data, and makes the best matching of the cloud data characteristics of the optical network to be scheduled according to the allocation mechanism of the preamble slot. Simulation results show that this model has better anti-noise performance for cloud data scheduling, higher matching degree of cloud data and higher scheduling accuracy.